CA2187792A1 - Complex combinatorial chemical libraries encoded with tags - Google Patents

Abstract

Encoded combinatorial chemistry is provided, where sequential synthetic schemes are recorded using organic molecules, which define choice of reactant, and stage, as the same or different bit of information. Various products can be produced in the multi-stage synthesis, such as oligomers and synthetic non-repetitive organic molecules. Conveniently, nested families of compounds can be employed as identifiers, where number and/or position of a substituent define the choice. Alternatively, detectable functionalities may be employed, such as radioisotopes, fluorescers, halogens, and the like, where presence and ratios of two different groups can be used to define stage or choice. Particularly, pluralities of identifiers may be used to provide a binary or higher code, so as to define a plurality of choices with only a few detachable tags. The particles may be screened for a characteristic of interest, particularly binding affinity, where the products may be detached from the particle or retained on the particle. The reaction history of the particles which are positive for the characteristic can be determined by the release of the tags and analysis to define the reaction history of the particle.

Description

2 1 877q2 COMPLEg '- 'TORIA]~ TJT~MTt'Z~r. T.TRR~l~TT.'!q ENCODED WIT~ TAGS
This Arpl;t~tlnn is a t-~ntin~lAt;~ in-part of U.S. Serial No. 08/227,007, filed April 13, 1994 the c~rtpntR of which are hereby incorporated by reference into the subject application .

TntroductiOn ~echnical Field The field of this invention f~r~rr~rnC I ' ;nAtt~rial chemistry which involves syntheses having a plurality of 10 stages, with each stage involving a plurality of choices, where large numbers of products having varying compositions are obtained.
Backcrround of the Invention 5 There is substantial interest in devising facile methods for the synthesis of large numbers of diverse _ ~ ds which can then be screened for various possible physiological or other activities. Typically such a synthesis involves successive stages, each of which 20 involves a chemical modification of the then existing molecule. For example, the chemical modification may involve the addition of a unit, e.g. a monomer or synthon, to a growing seguence or "l;f;~Ation of a functional group. By employing syntheses where the chemical 25 modification involves the addition of units, such as amino acids, nucleotides, sugars, lipids, or heterocyclic compounds where the units may be naturally-occurring, synthetic, or comb;nAt;~nc thereof, one may create a large number of cl __ ~c Thus, even if one restricted the 30 synthesis to naturally-occurring units or b~l;lrl;n~ blocks, the number of choices would be very large, 4 in the case .

of nucleotides, 2 0 in the case of the common amino acids, and essentially an unlimited number in the case of sugars.
One disadvantage heretofore inherent in the production of S large numbers of diverse, , flq, where at each stage of the synthesis there are a significant number of choices, is the fact that each individual compound will be present in a minute amount. While a characteristic of a particular, , fl, e.g. a physiological activity, may be 10 fl-t-rm;n-hle, it is usually impogsible to identify the chemical structure of the particular ~ o~nfl present.
Moreover, physiologically-active compounds have historically been discovered by assaying crude broths 15 using Edisonian or stochastic techniques, where only a relatively few ~ _ ~ul-ds are assayed at a time, or where a limited number of structurally similar homologs of naturally-occurring physiologically-active compounds are assayed. Two major problems have been associated with the 20 use of such crude broths, namely, the necessity to purify the reaction mixture into individual ~ --t ~-n~.rollnflA
and the time--nnq-~min~ effort required to establish the structure of the , _ ~ once purif ied .
25 To address these disadvantages and problems, techniques have been developed in which one adds individual units as part of a chemical synthesis se~-nt i -l l y, either in a controlled or a random manner, to produce all or a substantial proportion of the possible compounds which can 30 result from the different choices possible at each sequential stage in the synthesis. However, for these techniques to be successful it is n_.~ rry for the __ flq made by them to be amenable to methods which will allow one to fl--t-rml n_ the composition of a 35 particular compound so made which shows a characteristic of interest.

W0 95/28640 ~ 0 One such approach involves using a chip which allow3 for separate analysis at physically separate sites on the surface of the chip (Fodor et al., Science 251: 767 [1991]). By knowing which reactant is added gerl7/~nt;~11y at each such site, one can record the sequence of events and thus the series of ro~C~inn~. If one then subjects the chip to a screening method for a particular desired characteristic and detects the characteristic one can readily determine the _ i synthesized at the site which demonstrates that characteristic.
Another such terhn; ~-~ involves the theoretical synthesis of oli~nnl-r1 Pntides in parallel with the synthesis of oligopeptides as the c ~ ullds of interest ~Brenner and Lerner, PNAS USA [1992] 81: 5381-5383).
Further techniques are also disclosed in the following publications: Amoto, Science (1992) 257, 330-331, discusses the use of cosynthesized DNA labels to identify polypeptides. Lam, et al., Nature (1991) ~L, 82-84, describe a method for making large peptide libraries.
TTnllghtnn, et al., Nature (1991) 354, 84-86, and Jung and Beck-Sickinger, Angew. Chem . In t . Ed . Engl . ( 19 92 ) 91, 367-383, describe hnr~nlo~y for making large peptide libraries. Kerr et al., J. Amer. Chem. Soc., (1993) 115, 2529-31, teach a method of synthesizing oligomer libraries encoded by peptide chains. Finally, international applications WO 91/17823 and WO 92/09300 concern n~trrial libraries.
However, since methods such as the preceding typically require the addition of like moieties, there is substantial interest in discovering methods for producing compounds which are not limited to sequential Rfifl;t;on of like moieties. Such methods would find application, for example, in the modification of steroids, antibiotics, sugars, coenzymes, enzyme inhibitors, ligands and the 2l 87792 Wo 95n86,~0 like, which fres~uently involve a multi-stage synthesis in which one would wish to vary the reagents and/or conditions to provide a variety of compounds. In such methods the reagents may be organic or inorganic reagents, 5 where functionalities may be introduced or modified, side groups attached or = removed, rings opened or closed, stereochemistry changed, and the like. (See, for example, B~unin and Ellman, ,JACS 114, 10997 [1992] . ) For such a method to be viable, however, there needs to be a 10 convenient way to identify the structures of the large number of compounds which result f rom a wide variety of different modifications. Thus, there is a need to find a way whereby the reaction history may be recorded, and desirably, the structures of the resultant ~ _ ~A
i~nt; f; ~, ~
Finally, as the size of a library of compounds 80 synthesized increases, known tF~hn;sr~A of ~structure el~ t;nn and product segregation introduce substantial 20 inefficiencies and uncert~;nt;~oA which hinder the accurate determ; n~t i~n of the structure of any compound identif ied as being of interest. Thus, there is a sllhAt~nt;~l need for new methods which will permit the synthesis of complex c ' ;n~torial chemical libraries which readily permit 25 accurate structural determi~ation of individual compounds within the library which are i~l~nt; f ied as being of in terest .
Many of the disadvantages of the previously-described 30 methods as well as many of the needs not met by them are addressed by the present invention which, as described more fully hereinafter, provides myriad advantages over these ~revious1y-described methods.
.

~ wog~o 2 1 87792 ~ c~

S~ rY of The Inventio~
Methods and compositions are provided for encoded in~tQrial libraries, whereby at each stage of the synthesis, a support, such as a particle, upon which a 5 ~ ' is being synth.oc; 7~'d is uniquely tagged to define a particular event, usually chemical, associated with the synthesis of the compoLmd on the support. The tagging is ~r~~ l;ch~A uging ;-l~nt;fler molecules which record the sequential events to which the supporting particle is 10 exposed during synt~lesis, thus providing a reaction history for the compound produced on the support.
Each ;~ nt;f;~r ~l~c~ is characterized by being stable under the synthetic condit iong employed, by L~ ;n;n~
15 associated with the supports during the synthesi~, by uniquely riPf;n;n~ a particular event during the synthesis which ref lects a particular reaction choice at a given stage of the synthesis, by being distingll;chAhl~ from other ~ c that may be present during assaying, and 20 by allowing for ~f-t~ of a tag _ ~ which is disc/~rn;hll~ by a convenient, analytical te~-hniq.l~.
The ; ci~nt; f i F~r8 of this in-~rention are used in combination with one another to forn~ a ~inary or higher order 'n~~~;nS
25 system permitting a r~lativ~ly small number of ;~lPnt;f;ers to be used to encode a re' atively large number of reaction products. For example, when used in a binary code N
nt;f;ers can uniquely encode up to 2~ different _ _ ' flR, Moreover, the i fl~nt; f j ~rR of this invention need not be bound serially t_rough a previous i~ipnt;fi~r but rather are individually bound to the substrate, either directly or t~rough the product being synthegized. The i~lPnt;f;~rS
35 are not sequencable. Purthermore, the ;~nt;f;~rs contain a cleavable member or moiety which permits ~ t;~ of a tag ~ which can be readily analyzed.
_ .. ............. . . .. . _ _ _ _ _ . _ _ _ _ W0 95/286~0 2 1 8 7 7 9 2 P

Conveniently, the combinatorial synthesis employs llPf;n~hle solid gupports upon which rp~ct;nnc are performed and to which the i~lont;f;ers are bound. The individual solid supports or substrates carrying the final 5 product ~ ds may be screened for a characteristic of i~terest and the reaction history ~PtPnrn;nP~I by analyzing the ~oci ted iA~;f;~r tags.

.

~ =--wo 95l28640 2 1 8 7 7 9 2 1 ~., s ~
.

i3rief De3cri~tion of the Drawinqs Figure 1 illustrates the analysis of tag 4 by mass spectroscopy. Two signals corrP~p~n-l; n~ to tag 4 were observed .

Figure 2 illustrates the analysis of tag 11 by mass spectroscopy. Two signals CU11 ~uullding to tag 11 were observed .
10 Figure 3 illustrates the analysis of tag 13 by mass spectroscopy. Two signals corresponding to tag 13 were observed .
Figure 4 illustrates the analysis of tags 4, 11 and 13 by lS positive chemical ;on;7~t;on mass spectroscopy (PCIMS) when apprn~ tPly e~ual amounts of each tag were mixed together. Two signals ~ ullP~ l; n~ to each separate tag could easily be distinguished.
20Figure 5 illustrates the mass spectrum and L~ t~ram for the underivativzed d-7 sample only.
Figure 6 illustrates the mass spectra and c:l l, to~rams showing the ; _ u v ~ t in ,~ t s~aphy due to 25 derivatization of tags, a d-5 and d-7 mixed sample are "~E/P~8Pd .

DE~rATT Rn DESCRIPTION OF THE INVENTION
As used in this ~rpl; ~ ~tinn the term "tag" or "T" means a ~hPm;~l moiety which possesges two properties. First, it is capable of being distinguished from all other chemical moieties. Second, it ia capable of being detected when present at 10-1b to 10-9 mole. These two properties may be embodied in a single chemical structure. Alternatively, these properties may be embodied in separate chemical structures which are linked together. In this latter ca,se, one of the chemical structures, which may be designated C (or in the case of more than one such structure C, C', etc. ) provides the property of rendering the tag distinguishable from other tags while the other chemical structure, E, provides the property of rendering the tag ~.ot~nt~hl~ and optionally may provide the property Of rF.n~Pr;n~ the tag geparable from other tags.
AB used in this ~rpli~atinn, the term ~linker~ or "L"
means a chemical moiety which possesses three properties.
First, it is ~tt~rh~hlp to a solid support. Second, it is attachable to a tag. Third, when it is attached to both a solid support and a tag, it is cleavable such that the tag may be released from the solid support. These three properties may be: ` -';P-3 in a single chemical structure.
Alternatively, these properties are embodied in three chemical structures which are linked together. In this latter case one of the rhPm;~ L- uuLuL~:s, which may be designated F1, provides the ~lu~elLy of rendering the ~ ker attachable to the solid support; the second chemical structure, which may be designated V, provides the property of rendering the linker cleavable; and the third chemical structure which may be designed A', provides the property of rendering the linker ~tt~nh~hle to the tag. Desirably, the chemical structures V and A' are one and the same, in which case V-A' may be designated F' .

_9_ As used in this application, the term '~ pnt;f;~rll means a chemical entity which includes both a tag and a linker.
Thus, in the broadest sense an identif ier may be represented by the formula L-T while specific embodiments 5 of the ;~nt;f;er may be represented by the fu., l~o F1-V-A'-T; F1-V-A'-C-E (or Fl-V-A'-E-_); ~-C-E (or ~-E-C); and L-C-E-C' .
As used in this application, the term "bound identifier"
10 means an i~ont;f;er ~tt~hPd to a solid support.
As used herein, the term ~'choice" means the alternative variables for a given stage in a combinatorial synthesis, such as reactant, reagent, reaction conditione, and 15 combinations thereof . The term ~'stage" CUL ' ~ /u~-ds to a step in the sequential synthesis of a olln~l or ligand;
the compound or ligand being the f inal product of a ; n;~tt~rial synthesis .
20 The term "alkyl" includes linear, branched, and cyclic structures and combinations thereof. Thus, the term includes methyl, ethyl, propyl, iaopropyl, butyl, sec- and tert-butyl, cyclopropyl, cyclobutyl, cyclopentyl, 2-methylcyclopropyl, and the like. Lower alkyl is C1-C6 25 alkyl. ~ower alkenyl is C2-C6 alkenyl of a linear, branched, or cyclic configuration and combinations thereof .
Unless otherwise ;nrl;r~ted~ it is ;nt~-nrl~d that the 30 definitions of any substituent (e.a., R1 R', Z, etc. ) in a particular molecule be ;n~ r~n-l~n~ of its definitions elsewhere in the molecule. Thus, NR~R~ represents NHH, NHCH3, NHCH2CH3, N (CH3) " etc .
,^ .
35 Some of the, _ ~lq r~ ( r;h.-rl herein contain one or more centers of asymmetry and may thus give rise to enantiomers, diastereoi~ , and other steroisomeric Wo95128640 }~

forms. The pre6ent invention i8 meant to include all such possible stereoisomers as well as their racemic and optically pure forms. Optically active (R) and (S) isomers may be prepared using chiral synthons, chiral 5 reagents, or resolved using convPnt;nn~l terhn;~lPc. When the , ' - described herein contain olef inic double bonds, it is ;ntPn~lP~l to include both E and 2 geometric isomers .
10 The materials upon which the, ' ln~tnrial syntheses of this invention are perf ormed are ref erred to herein interchangeably as beads, solid surfaces, (solid) substrates, particles, supports, etc. These terms are intended to include:
a) solid supports such as beads, pellets, disks, capillaries, hollow fibers, needles, solid fibers, cellulose beads, pore-glass beads, silica gels, polystyrene beads optionally cross-linked with divinylhpn7~npl grafted co-poly beads, poly-acrylamide beads, latex beads, dimethylacrylamide beads optionally cross-linked with N,N'-bis-acryloyl ethylpnptl;Am;np~ glass particles coated with a hydrophobic polymer, etc ., i . e ., a material having a rigid or semi -rigid surface; and b) soluble supports such as low molecular weight non - cros 8 - linked po lystyrene .
These materials must contain functionalities or must be 30 able to be f~lnrt;nn~l;7ed such that ;~lPnt;C;ers or product ;ntP ~;;qtpc may be ;Itt~rhf~l to them.
In addition, the following abbreviations have the indicated meanings:
AcOH = acetic acid BSA - bis (trimethylsilyl) ~rf~t~
CAN = cerium (iv) ammonium nitrate wo gs/28c40 2 1 8 7 7 9 2 ~", ~
DEAD = diethyl azodicarboxylate DCM = dichloromethane DIC = diisopropylcarbodiimide DMF = N, N-dimethylformamide 5 Fmoc = 9 - f luorenylmethoxycarbonyl HOBt = l-llydL~yLellzotriazole PhMe = toluene r. t . = l room temperature TFA = trifluoroacetic acid THF = tetrahydrofuran The subject invention ~nnl-PrnC the production of libraries of products, i.e. ~ _ Ju--ds, where the individual products or, __ ~c present in the libraries may be physically separated from one another and may be screened for a characteristic of interest either bound to, or detached from, a solid support. By having serial syntheses, where at each stage of a 8ynthesis each of the individual int~ -'iAtoc i8 treated in a variety of ways, a very large number of products is produced, eac~ of which is present in a small amount, frequently less than 100 pmol, more frequently less than 10 nmol. Because of the small quantity of final product or ~- ~Ulli 80 produced, identifying these products by isolating and structurally Pll~r;~t;n~ the products would generally not be feasible.
Moreover, in sp~lpnt;~l synthesis involving other than the addition of similar units, the analysis would be arduous if not impossible using the amount of product typically available. However, by associating each choice or ' ;n~t;on of choices (e.g., "add reagent A" or "add reagent A, then reagent B, and heat to 100C for 2 hrs. " ) of the serial synthesis with a, ' ;n~tinn of i~lPnt;f;ers which def ine the choice of variable5 such as reactant, reagent, reaction conditions, or a c -n~t;nr~ of these, one can use the itlPnt;~;Prs to define the reaction history of each ~Pf; n~hl e and separable substrate . The analysis of tags ~Pt~hP~l from the ;dpnt;~;prs allows for ready _, _ W095/2864l0 ~ 'C

;r~,ont;f;cat;t n of the reaction higtory, at picomolar or lower concentrations, e . g. f emtomolar or less . One can determine a characteristic of a product of a synthesis, usuaily a chemical or biological characteristic by various 5 screening terhn;qn~, and then identify the reaction history and thereby the structure of that product, which has the desired chEracteristic, by virtue of the tags ~or; ~ted with the product .
10 The use of the ingtant mu'-iple tag system avoids the necessity of carrying ou~ a, ~ 1; r~ted cosynthesis which reduces yields and require~ multiple protecting groups, and avoids the n~r~cql ty of using se~lrnr~hle tags which are n~c~sslrily chemically l~bile. Both the necessity of 15 multiple protecting groups and the intrinsic instability of all known sequencable tagging molecules (i.e., nucleic acid or peptide oligomers) severely limit the chemistry which may be used in the synthesis of the library element or ligand.
Moreover, the use of a bi~ary, or higher order, multiple tag 6ystem reduces enormou81y the number of tag8 nf~rf~cFIslry to encode the reagent/reac~.~nt choice in any stage in a synthesis. For exampl~, if a particular synthetic stage 25 could be carried with 128 ~ifferent choices for reagent, the binary system would require only 7 tags. This can make the difference between a practical ~nro~;n~ system and an impractical one, because it may not be feasible to ob~ain and use the large number of distinguishable tags 30 required by other systems. With the binary system of the invention, 30 dist;n~l;~hSlhle tags are available and are suEficient to encode ~109 different syntheses.
Importantly, the present method employs tags which are 35 rlP~-~rh:~hl e from a ligand or golid support for the purpose of ~corl; n~ . Such ~t~rh;lh; l; ty also allows the tags to be distinguished on more than one basis; in particular, Wo 95/28640 P~ll~ t -~

they can be separated (e . q ., on the basis of ChLI tt~raphic retPn~inn time) and then analyzed (e.t~., a second basis is a spectral p, u~el Ly such as mass spectroscopy m/e, or electrophoricity). Having multiple 5 bases for distinction allows the Pnt~n~l;n~ of large amounts of information with a small number of tags.
Detat' - further allows tags to be detected at very low levels, because they can be removed from the support lO matrix on which the synthesis is effected and from the ligand synthesized, the presence of either of which could provide spurious background gignalg, e.g. by t~lPnt~h;ng f luorescence or the like .
15 DPta~h~hle tags are also ~hlP to rapid analygis by automated I l ;n~ systems, and allow for selective deriv~t; 7a~;t n for detection via functional groups, t~l ;m;n~;n~ any ;n~ ~;hility between the detection moiety and the reaction conditions used in the synthegis.
Inherent in any tagging scheme is the requirement that the chemical characteristics of the tags and the chemical stages f or their incorporation be ~ ; hlP with the characteristics of the ligand and the stages in their 25 synthesis, and vice versa. The advantage of tags that are generally unreactive, as ~ l; f i Pr9 hereinaf ter by the substituted-aryloxypolymethylene moieties, is a greater range of t~hPm;t-~1 transformations and chemical functionality that can be employed in synthesis of the 30 ligands.
A further advantage of the rhPm;t~l1y stable tags of this invention is their compatibility with a greater variety of rapid, convenient methods of separation and analysis, such 35 as gas ~1" o~raphy and mass spectrometry. ~ Ve~L, the organic tags of these invention generally do not specif ically interact with biological receptora . Thus, 21 877q2 Wo 9s/2864~

these tags will generally not give spurious results in biological assays and will generally not be modified by enzymes or other biological molecules.
5 Finally, the rh~m1r~1 stability of the present tags allows them to be detached by a wide variety of methods which improves aensitivity in their analysis as described above.
Thus, this invention provides methods and compositions for lO encoded ' ;n~tr,rial synthesis whereby at each stage of the synthesis one or more identifiers are provided which encode an event associated with a particle stage in the synthesis of a, _ ~ on a support or particle. This event comprises the choice of reactant and/or reaction 15 conditions at that stage of the reactions where each such stage may involve one or more r~ctAntR which are the same or different under the same or:different conditions, e.g.
partial reactions, multiple additions, rate of addition, differing, ' ;n~tirnR of reagentg, etc. In a~lrl;t;rn, 20 groups of particles may be sequestered from other groups of particles and subjected to a different series of events at any time during the course of the RPqn~nt;~1 synthesis.
By providing N identifiers, each having M disting~;Rh~h1e 25 states, M~ different syntheses can be uniquely defined. In the case of M=2 where the two atates could be the presence or absence of j~nt;f;~r, the synthesis would thus be defined by a base 2 or binary code. In the case of M=3 wllere the three states could be the presence of an 30 identifier at two disting~;Rh~h1e r.JI,r~ tions or its absence, the synthesis would be defined by a base 3 code.
~lerein, such base M codes where M~2 are termed higher order codes. The advantage of higher order codes over a binary code is that fewer itl~ntifirrR are required to 35 encode the same quantity of information about the synthesis. The products which are produced will be defined as resulting from a serial synthesis. }~t each . _ _ _ _ _ _ _ _ _ _ . . . ... .. , . . _ _ ..

stage in the synthesis, there is available a plurality of reactants and/or reagents and/or conditions, which result in a feature of the product in relation to an ;~ nt;f;~h~e and usually separable entity, e.g. tag. In referring to reactants and reagents, it i8 ;ntPn~F~i that the reactant, for the most part, becomes incorporated into the product, e.g. an amino acid, nucleotide, nucleophile, electrophile, diene, alkylating or acylating agent, diamine, or any other synthon, etc. while a reagent may or may not become incul~u~ ~Led into the product, e.g. base, acid, heat, n,r;ll;7;n~ or reducing agent, while both will be included under the term "agent". The synthesis may involve individual rP~'t~ntC which become incorporated into the product. Alternatively, a stage may involve one or more reactions which result in a modification of a reaction intermediate. In many cases, combinations of these pn~3R;h;l;ties Will be involved.
Using a base 2 or binary code (M=2) and three i~i~ont;fiers (N=3), as many as 8 (23) agents for a given stage in a synthesis may be encoded. If the three ; rl~nt; f; orS are represented as T1, T2, and T3 and the presence or absence of each i ~Pnt; f; er is represented as a ' 0 ' or ' 1~
respectively, then eight different agents could be 25 represented in a binary code as follows:
Agent 1 Agent 2 Agent 3 Agent 4 Tl,T2,T3 0,0,0 1,0,0 0,1,0 1,1,0 Agent 5 Agent 6 Agent 7 Agent 8 Tl,T2,T3 0,0,1 1,0,1 0,1,1 1,1,1 Similarly, even more inforr-t;nn about the synthesis may be erLcoded by mo~e if~f~nt;fi~rR. For example, 9 nt; f; P~5 ~N=3 ) and a base 2 code (M=2 ) would allow up to 29 or 512 different agent choice8 to be encoded. Using 35 a base 3 code (M=3) and three ;~nt;f;~rs (N=3) would _ _ 21 877~2 W0 9~28640 allow as many as 27 ~33) agent choices to be encoded. If tlle three ; ~ nt; f; ~rs are represented as T1, T2 and T3, and the absence of an; rl~nt ; f ; er i8 represented as a ' 0 ', its presence at a quantity of ~0.5 pmol/bead as a '1', and 5 its presence at a quantity at ~1 . o pmol/bead as a ' 2 ~, tllen the 27 different agents could be ~ es~l,Led by three nt; f; .~rs in base 3 code as:
Agent 1 Agent 2 Agent 3 Agent 4 Tl,T2,T3 0,0,0 1,0,0 2,0,0 0,1,0 Agent 5 Agent 6 . . . Agent 2 7 Tl,T2,T3 1,1,0 2,1,0 ... 2,2,2 To make such higher order encoding schemes practical, one 15 additional i~n~; f; .o~ at a given quantity (e . g ., ~1. O
pmol/bead) would be added to all members of the library to provide a standard against which the qn~nt;t;es of all ; ~nt; f; erg would be meagured. The ~lAnt i t; f"R of the ;~Pnt;f;ers could be measured by gas ~:l~, tography or 20 HPLC with a variety of rl~te~t;nn methods. In the case of HPLC, quantities could be conveAiently measured by 81~;nt;l1~t;~ ntin~ if the ;~l~nt;f;ers were r~l;o~rt;vely labeled by differeAt quantities of a r~ nll~ . guch as tritium (3H) . It would be particularly 25 convenient to carry out the quantitation by measuring the ~H-to-1iC ratio, thus using 1~C as a standard. IA this way, as many as ten quantities of 3H could be distinguished to create a base 10 or ~cimal code (M=10) which could encode e~ol, ~ amounts of information with very few i~nt;f;ers.

3~

wo 95n~^640 2 l 8 7 7 9 2 , ~

Products and SYnthetic Stratec~ie3 For the most part, the products of the method of this invention will be organic , r~ln iC where the serial synthesis will involve the addition or removal of chemical 5 units, reactions involving the modif ication or introduction of one or more functionalities, ring openings, ring ~ los;ngc, etc. Chemical units can take many forms, both naturally-occurring and synthetic, such as nl~~leor~h;l~oR~ electrophiles, dienes, alkylating or 10 acylating agents, rll~m;n~c, nucleotides, amino acids, sugars, lipids, or derivatives thereof, organic monomers, synthons, and combinations thereof. Alternatively, reactions may be involved which result in alkylation, acylation, nitration, halogenation, nY;~3Ation~ reduction, 15 hydrolysis, substitution, .~l ;m;n~t;~n~ addition, and the like. This procegs can produce non-nl;s ;,, ~l;~; - ~, or ~- ' ;nAt;nnc thereof in extremely small amounts, where the reaction history, and composition in appropriate cases, can be defined by the present tags. Non-oligomers 20 include a wide variety of organic molecules, e.g.
heterocyclics, a~ c, alicyclics, aliphatics and c, '-;n~tinnc thereof, comprising steroids, antibiotics, enzyme inhibitors, ligands, ho AC~ drugs, alkaloids, opioids, terpenes, porphyrins, toxins, catalysts, as well 25 as _ ;n~t;nna thereof. Oligomers include oligopeptides, oligonucleotides, oligosaccharides, polylipida, polyesters, polyamides, polyurethanes, polyureas, polyethers, poly (~hns~hnrus derivatives) e.g. rh^~rh~tes, rhnSrhnni~t~a, rhnSrh~ .R, riln~h./.. ~.. ~ C, phosphites, 30 rh~ srh;n~m;de5, etc., poly (sulfur derivatives) e.g.
sulfones, sulfonates, sulfites, sulfonamides, sulff~n~m;~ a~ etc., where for the rhnsrhnrous and sulfur derivatives the indicated heteroatom for the most part will be bonded to C, H, N, O or S, and combinations 3 5 thereof .

w0 95128640 ;nnA may involve r ';~;r~tlnnc at a vari~ty of random sites of a central core molecular structure or modifications at a specific site. For example, one may LL~ n~te a polycyclic compound, where LL~ n~t;nn may 5 occur at a plurality of sites or use a LL- 'n:~tlng agent which will be specific for a particular site, e.g., N-bromos~rr;n;m;~e. For the most part, r~rt;nnC will involve single sites or equivalent sites, for example, one of two hydroxyl groups of a glycol.

For the most part, the subject synthesis will have at least two stages where other than bifunctional compounds are attached using the same linking fllnrtinn~l;ty, e.g.
amino acids and amide bonds, nucleotides and phosphate 15 ester bonds, or mimetic, 'q thereof, e.g., aminoiso-cyanateE and urea bonds.
The methods of the invention permit variation in reaction at each stage, depending on the choice of agents and 20 conditions involved. Thus, for amino acids, one may have up to 20 amino acids involved using the common naturally-encoded amino acids and a much wider choice, if one wishes to use other amino acids, such as D-amino acids, amino acids havi~g the amino group at other than the ~-position, 25 amino acids having different substituents on the side chain or substituents on the amino group, and the like.
For the different nucleic acids, there will usually be up to 4 natural nucleic acids used for either DNA or RNA and a much larger number is available if one does not choose 30 to uEe those particular nucleic acidE. Por the sugarE and lipids, there are a very large number of dif f erent compounds, which . , '~ may be further increased by various substitutions, where all of these - ~lq may be used in the synthesis. For individual organic r~ q 35 the choice may be astrn~l rAlly large. In ~;tinn, one may have mimetic analogs, where ureas, urethanes, carbonylmethylene groups, and the like may substitute for _ _ _ _ _ , . , . . _ . . _ ... . . _ _ _ _ _ _ WO 95~28640 the peptide linkage; various organic and inorganic groups may substitute for the phosphate linkage; and nitrogen or sulfur may substitute for oxygen in an ether linkage or vice versa.
The synthetic strategies will vary with the nature of the group of products one wishes to produce. Thus, the strategy must take into consideration the ability to stage-wise change the nat.ure of the product, while 10 allowing for r~ot~nt;nn of _he results of the previous stages and anticipatin~ needs for the future stages.
Where the various units are of the same family, such as nucleotides, amino acids and sugars, the synthetic strategies are relatively well-est~hl;Rh~ and frer~uently 15 conv~nt;nn~l chemistry will be available. Thus, for nucleotides, pho .~hu,c.,,~idite or phosphite chemistries may be employed; for oligopeptides, Fmoc or Boc chemistries may be employed where conv~nt;nn~l protective groups are used; for sugars, the strategies may be less conv~nt;nn~l, 20 but a large number of protective groups, reactive functionalities, and conditions have been est~hl; Rh~rl for the synthesis of polyRacrh~ri-l~R. For other types of chemistries, one will look cr, the nature of the individual unit and either synthetic ~,J~o~Lu.lities will be known or 25 will be devised, as appropriate.
In some instances, one may wish to have the same or dif f erent blocks introduced at the same or dif f erent stages. For example, one may wish to have a common 30 peptide f~nrt;nnAl unit, e.g. the fibronectin binding unit (RGDS), a polysaccharide, e.g. LeX, an organic group, e.g.
a lactam, lactone, benzene ring, olefin, glycol, th;oeth~r, etc. introduced during the synthesis. In this manner one may achieve a r - ler~ r context into which the 35 variation is introduced. These situations may involve only a f ew stages having the plurality of choices, where a large numher of products are produced in relation to a 21 877~2 wo ssns640 particular functional entity. This could have particular application where one is interested in a large number of derivatives related to a core molecule or unit known to have a characteristic of interest.
In developing synthetic strategies, one can provide f or batch synthesis of a few compounds which would be prepared during the course of the ' ;n~tnrial synthesis. By taking extreme examples, for~ example, syntheses which 10 might involve steric h;n.1r,~nrG, charge and/or dipole interactions, alternative reaction pathways, or the like, one can optimize cnn~;tinn~ to provide ~or ~nh~n-l~d yields of compounds which might not otherwise be formed or be formed only in low yield. In this manner, one may allow 15 f or a variety of reaction conditions during the combinatorial synthesis, involving differences in solvent, temperatures, times, concentrations, and the like.
Furthermore, one may use the batch syntheses, which will provide much higher rnncl~ntrr~t;cm~ of particular products 20 than the ' ;n~tnrial synthesis, to develop assays to characterize the activity of the ~ _ Ju..ds.
SuD~orts: Atta ' ~ ~n~l Detachment The synthetic protocol retauiæs that one provide for a 25 plurality of different r~rt~nnr involving different r~ rt~nt~ resulting in a plurality of different ;nt, ~;~t.,5 at each stage of the synthesis. While other techniques are available, this can be achieved most conveniently by employing small ~ f; n~hl e solid 30 substrates, commercially available as beads, which can be readily mixed, separated, and serve as a solid substrate ~or the seS~uential synthesis. The solid substrates may be solid, porous, deformable or hard, and have any convenient structure and shape. In some instances, magnetic or 35 fluorescent beads may be useful. The beads will generally be at least 10-2000 llm, usually at least 20-500 llm, more usually at least 50-250 llm in diameter.

W0 ~5128640 1 _ I/L_ Any convenient composition can be used for the particles or beads, which bead composition will r-;nt~;n its mechanical integrity during the various proces6 stages, can be f-lnrt;nn~l;7t~tl, has functional groups or allows for reaction with an active specie3, allows for the serial synthesis as well as attachment of the ;tlt~nt;fiers, can be readily mixed and separated, and will allow for convenient det:-t' ' of the tags and products. Beads which may be employed include cellulose beads, controlled-pore glass beads, silica gel, pOly~LyLe:.-e beads, particularly polystyrene beads cross-linked with divinylbenzene, graf ted co -polymer beads such as polyethyleneglycol/polystyrene, polyacrylamide beads, latex beads, dimethylacrylamide beads, particularly cross-linked with N,N'-bis-acryloyl ethylene diamine and comprising N-t-butoxycarbonyl-~-alanyl-N'-acryloyl hexamethylene diamine, composites, such as glass particles coated with a hydrophobic polymer such as cross-linked polystyrene or a f luorinated ethylene polymer to which is grafted linear polystyrene; and the like. General reviews of useful solid supports (particles) that include a covalently-linked reactive fllnrt;rn~l;ty may be found in Atherton, et al., Pros~ectives in Pe~tide Chemistrv, Karger, 101-117 (1981); Amamath, et al., Chem. ~ev.
77:183-217 (1977); and Fridkin, The Pe~tides, Vol. 2, Chapter 3 , ~r~t~t~m; r Press, Inc., (1979), pp. 333-363 .
Depending upon the nature of the synthetic procedure or the assay of the final product, one or another bead may be more or less desirable. While beads are especially convenient, other solid supports may also f ind use, such as capillaries, hollow fibers, needles, solid fibers, etc., where the size of the solid support allows for the desired variation in reaction histories.
Depending upon the nature of the synthesis, the beads may be fllnrt;rn~l;7ed in a variety of ways to allow for Wo95/28640 2 1 87792 r ~

att~ of the initial reactant. These may be linked through a non-labile linkage such as an ester bond, amide bond, amine bond, ether bo~d, or through a sulfur, silicon, or carbon atom, ~lorPnrl;n~ upon whether one wishes 5 to be able to remove the product from the bead.
Conveniently, the bond to the bead may be pe~ allellL, but a linker between the bead and the product may be provided which is cleavable such as exemplified in Table 1. Two or more different 1 ;nk~Pq may be employed to allow for 10 differential release of tags and/or products.
T)PrPn~; n~ upon the nature of the linking group bound to the particle, reactive il-n~ti~n~l;t;Pc on the bead may not be np~pqc~ry where the manner of linking allows for 15 insertion into single or double bonds, such as is available with r~rhPnPq and nitrenes or other highly-reactive species. In this case, the cleavable linkage will be provided in the linking group which j oins the product or the tag to the bead.
Desirably, when the product is pPrr-n~ntly attached, the link to the bead will be P~rtPnrlPri, 80 that the bead will not sterically interfere with the binding of the product during screening. Various links may be employed, 25 particular hydrophilic links, such as polyethyleneoxy, saccharide, polyol, esters, amides, ~_ ' in~t;nnq thereof, and the like.
Functionalities present on the bead may include hydroxy, 30 carboxy, ;m;n~h~ P, amino, thio, active halogen (Cl or Br) or ~sp~ lh~lo~pn (e.g., -CF3, -CN, etc.), carbonyl, si 1yl, tosyl, mesylates, brosylates, triflates or the like. In selecting the functionality, some ~nqi~lpration should be given to the fact that the ;~Pnt;f;ers will 35 usually also become bound to the bead. ~nqi~lPration will include whether the same or a different functionality should be associated with the product and the i~lpnt;f;pr~
.

21 ~77~2 Wo95n8640 ~ 5"~

as well as whether the two functionalities will be , -t;hle with the product or ;APnt;f;er att~r~ t and tag ~lPt~ stages, as appropriate Different linking groups may be employed for the product, so that a specific 5 quantity of the product may be selectively released. In some instances the particle may have protected functionalities which may be partially or wholly deprotected prior to each stage, and in the latter case, reprotected. For example, amino may be protected with a 10 r~rhrhPnzoxy group as in polypeptide synthesis, hydroxy with a benzyl ether, etc.
Where ~lPtarl of the product is desired, there are numerous functionalities and rp~rt~ntR which may be used.
15 Conveniently, ethers may be used, where substituted benzyl ether or derivatives thereof, e . g . benzhydryl ether, indanyl ether, etc. may be cleaved by acidic or mild reductive conditions. Alternatively, one may employ ~_Pl;m;n~t;nn, where a mild base may serve to release the 20 product. Acetals, ;nrlllA;nrJ the thio analogs thereof, may be employed, where mild acid, particularly in the presence of a capturing carbonyl I ~ ', may serve. sy, ~ n;nr, formaldehyde, HCl and an alcohol moiety, an a-chloroether is f ormed . This may then be coupled with an hydroxy 25 fl~nrtir,n~l; ty on the bead to form the acetal . Various photolabile ~; nk~ R may be employed, such as o-nitrobenzyl, 7-nitroindanyl, 2-nitrobenzhydryl ethers or esters, ètc. Esters and amides may serve as linkers, where half-acid esters or amides are formed, particularly 30 with cyclic anhydrides, followed by reaction with hydroxyl or amino f~nrt;nn~l;ties on the bead, using a coupling agent such as a r~rho~;;m;~P. Peptides may be used as linkers, where the sequence is subj ect to enzymatic hydrolysis, particularly where the enzyme recognizes a 35 specific sequence. Carbonates and carbamates may be prepared using carbonic acid derivatives, e . g rh~sri~.nP, carbonyl diimidazole, etc. and a mild base. The link may Wo 95/28640 2 1 8 7 7 9 2 "~

be cleaved using acid, ba3e or a strong reductant, ~.g., LiAl~, particularly for the rA~hrnAt~ ester3. For a ~ist of cleavable linkages, see, for example, Greene and Wuts, Protective Group~ in Organic Synthesis, 2nd ed Wiley, 5 1991. The versatility of the various systems that have been developed allows f or broad variation in the conditions for att~ of products and ;~l~nt;f;ers aAd differential ~F't5 ' t of products and tags, as desired.
10 Tlle following table indicates various illustrative linking Ullit3 ~ e., F2 in E'ormula I) and the manner in which they may be ~l~ave~:

Wo95128640 2 1 8 7 7 92 ~", . ~

Table 1. Various illustrative linking units and the manner in which they may be cleaved.
5 Linking Group Cleav~ge ~eagent silyl fluoride or acid A hv B Ce(NH~)2(NO3)6 -NCO2 (L) ~ OH-, H', or LiAlH;
C 03, OsO~/IO~-, or KMnO~
D 1 ) 2 or Br~, MeOH
2 ) H30 ' -Si-(L) n~ t;nn, H', Br2, C12, etc.

G F- or H' H base, OH-x = keto, ester, amide, NO2, sulfide, sul~oxide, sulfone, and related electron withdrawing groups H3O' or reduction (e.g.
Li /NH3 ) J (~3P) 3RhCl (H) K Li,Mg, or BuLi M Hg N Zn or Mg x = halogen or 3 0 pseudohalogen O nl6;~1~tinn (e.g. Pb(OAc) or HsIo6) P base x = electron withdrawing group ' (L) shows the point of att~ ' ' of the tag or product .
_ _ _ _ WO95128640 21 87792 r~ J 5'114'"'~ --~2 _ ,~CH,O(L) OR o21j~CH20~L) 01~
E = ,~O(L) oQ RO~O(L) = ~(L) ~ (L) ~(L) _~ 0 (L ) Q XR
R \~~CR
o G = --5~ (L) = ~\~O~(L) 21 877q2 wo95n8640 -27- P~
I = ~~ \(L) J ~ /~/\O(L) 3r K = ~O--(L) - / 5~0~ (L
N = /~/O~
0~
= ~O~(L) O~ o~ 'L
p /~\X ~X OR ~\X
(L) W0 9~/28640 2 1 8 7 7 9 2 (1.) is the tag or product either directly bonded to the ln-l;rat-~ atom or indirectly bonded through a linking group such as C~(0)0, which~lirking group may provide a convenient functionality.
5 R is ~I or lower alkyl.
Linker .
The choice of linker for: the ligand will be part of the synthetic strategy, since the linking group may result in 10 a reGidual fllnrt;r~n~l;ty on the product. It is feasible to further modify the product after ~Pt::ll' ' from the bead. In designing the synthetic strategy, one can u5e a fllnrtirn~lity to be rf~ ;n~d in the product as the point of att~ ` t for the linking group. Alternatively, when 15 permitted by the nature of the product, one could use a cleavage or ,l~t~ method which removes the linking iunctionality, e.g., an arylthioether or silyl with a metal hydride or acid. Si~ce in many cases the synthetic strategy will be able to include a funct;~n~ e~l Clte for 20 linking, the functionality can be taken advantage of in choosing the linking group. In some instances it may be desirable to have different functionalities at the site of linking the product to the support, which may necessitate ~lsing different modes of linking, which modea must 25 ~rr ' te either the same ~t~ method or different det~ methods which may be carried out concurrently or r~n~erut;vely, e.g., irradiation with light and acid hydrolysis .
30 Of particular interest for hi~ding the i~ont;f;ers to the particle are r~rh~n~R and nitrenes which can insert between a carbon and hydrogen atom to form a covalent bond, or into an olefinic bond to form a cyclopropane (in the case of carbene) or an = aziridi~e (in the case of 35 llitrene).

With carbene or nitrene linking groups various substituted benzenes may be used, where the benzene is substituted with a group capable of providing a carbene: CHN2, COC~N2, SO,CHN2; or nitrene: N3, NO2, NO, SO2N3. The cArbt~nt~R may t 5 be generated from t~lA7oAlkAn~ derivatives by photolysis, thermolysis, or by treatment with low valent transition metal species, e.g., Rh(OAc)2. The nitrene may be generated by photolysis or thermolysis from azides; and from nltro, nitroso and azides by using tervalent 0 pht-~sFhoru5 I ~ R or low valent transition metals.
A t3roup of linker moieties (F1-F2- ) o~ particular interest include 2-nitro-4-carboxybenzyloxy, 2-nitro-4-t1; A7oAt-etylbenzyloxy, 4 or 5 A~i t~ thylcarbonyl-2-methu~y~he.lo.-y, and 2-methoxy-4, or 5-~ic,LLu~y~henoxy moieties .
Illustrative ~ ~q where T Le~Let~e,-Ls the tag, Z
represents a carbene or nitrene precursor or a carboxy group, and R is H or lower alkyl are as follows. For photochemical tag detachment (e.g., with ultraviolet light at about 350 nm): T 3-Z-2-nitrobenzyl ether, T 4-Z-2-nitrobenzyl ether, T 5-Z-2-nitrobenzyl ether, T 6-Z-2-nitLube,~yl ether, T 2-Z-4-nitrobenzyl ether, T 3-Z-4-nitrobenzyl ether, T 3-Z-2-nitrobenzyl carbonate, T 4-Z-2-nitrobenzyl rArht~nAte~ T 5-Z-2-nitrobenzyl carbonate, T 6-Z-2-nitrobenzyl carbonate, T 2-Z-4-nitrobenzyl carbonate, and T 3-Z-4-niLLuben~yl t~Arht~ni~te. For oxidative t~tA~' (e.g., using ceric ammonium nitrate): l-OT-2-OR-3-Z-benzene, l-OT-2-OR-4-Z-benzene, l-OT-2-OR-5-Z-benzene, l-OT-2-OR-6-Z-benzene, l-OT-4-OR-2-Z-benzene, and l-OT-4-OR-3-Z-benzene. For reductive or alkylative detAt-l -nt (e.g. with lithium/ammonia or methyl iodide):
T (2-Z-phenyl) thioether, T (3-Z-phenyl) thioether, and T
(4-Z-phenyl)thioether. For desilylative detAt t (e.g., using tetrabutyl i f luoride or acid): T dialkyl -(2-Z-phenyl) silyl ether, T dialkyl- (3-Z-phenyl) silyl W0 95/28641~ 2 1 8 7 7 9 2 ether, T dialkyl- (4-Z-phenyl) 8ilyl ether, T-dialkyl- (2-Z-phenyl) silane, T-dialkyl- (3-Z-phenyl) silane, and T-dialkyl- (4-Z-phenyl) silane.
~ ;n~tn~i~l Svnthesis _ The synthesis will usually involve stages involving at least 2 choices, f requently at least 4 choices, and may involve 10 choices or more. Generally, the number of choices per stage will not exceed about 100, more usually not exceed about 50. The number of stages will usually be at least about 3, more usually at least about 4, fres~uently at least 5, and not more than about 30, more usually not more than about 25, preferably not more than about 20, more preferably not more than about 10, frequently not more than about 8.
The number of choices and stages will usually result in at least a number of ~ which allows f or a suf f icient variety to provide a reasonable l; k/~l; hnod that at least one ~ _ ' will have the characteristic of interest.
The subject r thn~lnlo3y allows for producing greater than 25,000 ~ , usually greater than 50,000 compounds, preferably greater than 200,000 compounds, and a million or more may be produced. This will usually mean at least 20 ~ but may be 106 or more.
In some syntheses, a stage may only involve one or two choices, but this situation will usually be limited in relation to the Ilumber of c __ '~ one wishes to produce 3 0 and the particular synthetic strategy . In many of the strategies, the restricted number of choices, i.e., fewer than 5 choices, more usually 2 or fewer choices, will be limited to the greater of 40~ of the total number of s~tages or about 2 stages in the sP~l~nt;~l synthesis, more 35 usually limited to 20~6 of the total number of stages.

Reaction Procedure.
In carrying out the synthesis, one may initially begin with a number of beads, usually at least 103, more usually 10~, and desirably at least 105, while generally not 5 PYrPP~;n5 lolS/ more usually not PYrPP~;ng at least 101~.
DPrPn~lin~ upon the number of choices in the first stage, one will divide up the particles accordingly into as many cnnt~;ners. One can use microtiter well plates, individual rnn~;nPrs, columns, gels, Terasaki plates, 10 flasks, Merrifield synthesis vessels, etc. The particles will usually be divided up into groups of at least one particle each, usually a plurality of particles, generally 1000 or more, and may be 105 or more depending on the total number o~ particles and choices involved in the stage.
One would then add the appropriate agents to each of the individual r nnt~;nPrs to process them in stages and add the i~ipn~;f;ers which encode the reagent and stage. Each stage would provide the desired reaction. Once the 20 reaction (s) is complete, one may wish to wash the beads free of any reagent, followed by combining all of the beads into a single mixture and then separating the beads according to the number of choices for the next stage.
This IJLOCC:dLI'~ of dividing beads, followed by the tagging 25 and synthesis stages (or vice verga), and then rP: ~ ;n;n~
beads is iterated until the ~ ' n~tnrial synthesis is completed .
In some instances, the same reaction may be carried out in 30 2 or more rnnt~;nPrs to enhance the proportion of product having a particular reaction at a particular stage as compared to the other choices. In other instances, one or more of the stages may involve a portion of the beads being set aside and undergoing no reaction, 80 as to 35 enhance the variability associated with the final product.
In other situations, batches may be taken along different synthetic p~ y~.

2~ 877q2 wo ssns640 ~ Jv, 'C ~

In order to record or encode the synthesis history on the beads, at each stage one would tag the beads associated with each choice and stage with their own unif~ue con~bination of ;8~nt;f;ers ~lt~nAt~ly one may use a single tag to record or encode this synthesis history.
D~r~n~';n~ on the chemistries involved, this tagging may be done prior to, after, or r~nnl ;tc7ntly with the reactions which comprise each choice. Further, as a control, sample beads may be picked at any stage and a portion of their tags cleaved off and decoded to verify that the correct tags are bound to the sample beads.
A8 indicated previously, in gome instances, portions of the particles will be segregated into subsets, where each of the subsets would then undergo a different reaction series. At any time, the portions may be re~ ' ;n~f~. into a single mixture for subsequent reaction. For example, if at one stage one introduces unsaturation, one could provide two subsets, where in one subset the unsaturation is reduced, while in the other subset the unsaturation is epoxidized. These two subsets could then be subjected to dif f erent reaction series .
After synthesis of the products is complete, they are screened for a desired yL~,y~, Ly either after ~7,~t,R~7- ~ of the ligand from the bead or while still attached. In the latter case, beads, for example, may be incubated in aqueous buffer with mouse 1 --lon-~l antibody Y. After incubation and washing, the beads are ; n~l~h~tPd with i711r~71;np ~hnsph,7t ~Re-conjugated rabbit (or goat) polyclonal antibody directed against mouse ;~nt;hf~';es.
~3sing a f luorescent precipitation developing reagent, fluorescent beads with attached monoclonal antibody are ;r7~nt;f;ed and manually separated from the majority of clear, unstained beads. Alternatively, the fluorescent beads can be separated using a fluorescence-activated cell sorter, BO long as the tags are retained on the bead under .

W0 95~8640 ' ` ~

the conditions of sorting. Each selected fluorescent bead is sub; ected to a means f or releasing at least some portion of all of the tags f rom the bead .
5 In instances where the synthesis doe6 not involve the stagewise addition of like units, or where reaction byproducts are formed, there may be instances where there will be a plurality of ~ ~ on a single bead or the structure of the active UUl.ll~UUlld cannot be known from its 10 reaction history. In ~ ccordance with the subject invention, by knowing the ~eaction history, one may repeat the synthesis on a larger scale 80 as to obtain a sufficient amount of the product(s) to isolate the product (8) and structurally identify the active compound.
The sub; ect methodology may be illustrated using various reaction sP~ on~oR For example, barbiturates may be prepared by ` ;nin~ an aldehyde or ketone with an acetate ester to prepare a crotonate under Claisen 20 conditions to provide an unsubstituted to tetrasubstituted crotonate. The crotonate may then be, ' ;nf~ with a second acetate under Michael conditions, whereby a glutarate may be r~h~tA;n~d having up to 6 substituents.
The glutarate may t~.~n Le, ' ;n~l with ammonia or 25 monosubstituted amin6 ~o provide the barbiturate. By varying the aldehydes ~nd ketones, the A~et~t~ and the amines, a great variety of barbiturates may be obtained.
Where functionalities are present on one or more of the substituents, such as amino, carboxy, hydroxy, thiol, and 30 the like, these groups may be protected or modified as des ired .
In another example described by Bunin and Ellman, J. Am.
Chem. Soc., 114, 10997 (1992), bPn~ ;A7epines are 35 produced. One begins the synthesis with different amino protected substituted 2- n~h~n7Qrl~n-~n~ bound to individual particles through, for example, a 4'-oxy group.

Wo95/28640 21 87792 1~"1 /Q4 ~ --To each different group of particles in different vessels, after deprotection, are added a different Fmoc-protected cY-amino acid, either naturally occurring or synthetic, under conditions where a peptide bond is formed. After 5 deprotection, ;nt~.rn~l Cyrl;7~t;nn is caused, followed by alkylation on nitrogen with an alkylating agent. In only three stages, a very large number of b~n7o~ 7~rin~a are prepared and the libraries may be screened f or tr~n'I'~;1; 7 inj or other activity.
A wide variety of drug analogs may be produced, such as analogs of antihypertensive agents, e.g. enalapril;
,l~-blockers, e.g. prop~nolol; antiulcer drugs (~,-receptor Ant~rnn; ~ts) e.g. cimetidine and ranitidi~e; antifungal 15 agents (cholesterol-demethylase inhibitors) e.g.
; aor^n;q7~1P; anxiolytics, e.g. diazepam; analgesics, e.g.
aspirin, rh~nArrt~m;tl~, and fentanyl; antibiotics, e.g.
vancomycin, penicillin and cephalosporin;
Int ;; n f 1 ~ories , e . g . cortisone ; contraceptives , e . g .
20 progestins; abortifacients, e.g. RU-456; antihistamines, e.g. chlorrh.onAm;n~; antitussives, e.g. codeine;
sedatives, e . g. barbitol; etc .
An illustrative synthesis of r; t;~in~ analogs could 25 involve 1:YdL~ hylsubstituted histidines, and related h~terocycles, where the L~ ;ninrj carbon atoms or nitrogen atoms could be ~urther substituted or unsubstituted, ~ ?-Am;nn;~7kylthiols, and subatituted th;o~m;~;n~ esters, where the groups on nitrogen could be varied, such as 30 nitro, cyano, hydroxy, alkyl, combinations thereo~, and the like.
Ide~tifier The i~lrnt;f;rrs of this invention may be represented by 35 the Formula I:
F1-F2-C-E-C' ~ I

21 877q2 WO 95/28640 1._11L., C .

where Fl-F2 is a linker which allows for atta~ l t to a support and det~t' of the tag from a support; and C-E-C'is the tag which is capable of detection and disting~ h~h; l; ty;
E is a tag _ ^nt which (a) allows for detection, such as an electrophoric group which can be analyzed by gas ~ to~raphy or mass spectroscopy or (b) allows for detection and for separation;
C and C' are tag ,~ nPnt~ which allow for individually distinguishing one tag from all other tags, usually allowing for separation as a result of variable length or substitution, for example, varying the ~:h, ~ tographic r~t~nt;~n time or the mass spectroscopy ratio m/e;
F2 is a linking r _ ' capable of being selectively cleaved to release the tag c ,. l~ellt; and Fl i8 a functional group which provided for att~
to the support; or F2 is a bond when Fl is a cleavable group such as O~I
2 0 or carboxy .
Although the identif iers of Formula I are typically added at each appropriate stage and choice during the combinatorial synthesis, the portion E can be added at the end of the syntheses either before or after cleavage (preferably photochemically or oxidatively) from the substrate. Specifically, where C ~nt~;n~ OH, llHR4, or SE~, E can be attached to C prior to cleavage. Alternatively, if E is attached after cleavage, the point of att~' t at C may be where F2 was attached. This is ~ _l;f;ed in the scheme on the following page:

wogs/28640 2187792 P l/L '''I

~12 , ~ CU~ C o-~
O ~0 <~CHO H~(CRg)n--H
502Cl F~ F ~$
COCl N(CH3 )2 F 0 503--~CH2)n--H
~0--(CR~z)n H
F ~CH3)2 where S = substrate and n = 1-40 Z~tt~' t of the identifier to the substrate can be represented as follows:
Fl-F2-C-E-C' + S ----------~ S-Fl'-F2-C-E-Ç' where Fl'-F2-_-E-C' represents the i~Pnt;f;P~ residue attached to the substrate. For example, when the bead i9 flll,r~t;~n~l; 7eC~ with an ~mi- -hyl group and Fl i8 CO~X, then Fl' i5 -C(O)-; when the bead t-~nt~;n~ an unsaturated 30 bolld and Fl is N2CX-C(O)-, then Fl' is ~CEI-C(O)- or --CX2-C (O) - .
Of particular interest f or use as i ~Pnt; f; ers are n~ of Formula I of the Formula Ia:
Fl-F2- (C (E-~' ) .) b Ia W09~/28640 r~
wherein:
Fl is CO2H, CH2X, NRlRl, C (O) Rl, OH, CHN2, SH, C (O) CHN2, S (0) 2Cl, S (0) 2CHN2, N3, NO" NO, S (0) 2N3, OC (O) X, C (O) X, NCO, or NCS;
5 F2 i~i ~ ~ ~ --Si(R )2A-- , --Si(R )z-- , --051[R )z-- ~ c ~-- A
--N(R )C(o)o--,--CR =CR--~CRz)2-- , CR--CR--C(R )z--.
--C(R )z--CR--CR--~ --C(RI) A-- --O--C(R~) A--¦ O~<R _ Si(C33 ~(C R 21~ A-- ,--R~ (C R Z )2 A
~f A . --CR CR--CIR )z--A--, ~B;
--S--C (R )A-- , I~C~I2A-- , /~C~I, A--R O~`CJI2A-- , and ~`C~2A-- ;

WO 95/28640 2 1 8 7 7 9 2 ~ c ~

--S--C~R k A ~ --C(X)R--C~R )Z A--, --CtOH)R--C (R )zA--. --C(O~I)R--C(C~I2X)R--~ .
--C(O~)R--C(R )z--C(X)R--, --C(OR)(C~12C~lzX~-- .
A-- ORl ~ A--" ~ A-- ~OR
OR
I

A-- ~ A--A-- A--OR R
--o J~X Rl ' ~ O~
R~l --O~OR

Wo 95/28640 2 1 8 7 7 9 2 1 ~I/.J~

A is -O-, -OC (O) O-, -OC (O) -, or -NHC ~O) -;
C is a bond; Cl-C20 alkylene optionally substituted by 1-40 F, Cl, Br, C1-C6 alkoxy, NR~Ri, oR4, or ~HRi, or - [ (C (R4) 2) ~-Y-Z-Y- (C (R4) 2) =Y-Z-Y] p-; with the proviso that the maximum number of carbon atoms in C+C' i8 pref erab ly 2 0;
c~ is H; F; Cl; Cl-C20 alkyl optionally substituted by 1-40 F, Cl, Br, Cl-C6 alkoxy, NR4R4, oR4, or NHRi, or - [ (C (R4) 2) ~-Y-Z-Y- (C (Ri) 2) =Y-Z-Y] p-;
E is Cl-Cl0 alkyl substituted by 1-20 F, Cl or Br; or Q-aryl wherein the aryl is substituted by 1-7 F, Cl, NO2, SO2Rs, or subgtituted phenyl wherein the substituent is 1-5 F, Cl, NO2, or SO2Rs;
E-C' may be -H, -OH, or amino;
15 Rl is H or Cl-C6 alkyl;
R3 is C=O, C(O)O, C(O)NRl, S, SO, or SO2;
R' is H or Cl-C6 alkyl;
Rs is Cl-C6 alkyl;
a is 1-5;
20 b is 1-3;
m and n is each 0-20;
p is 1-7;
Q is a bond, O, S, NHR4, C=O, -C(O)NRs, -NRsC(O)-, -C(O)O-, or -OC (O) -;
25 X is a leaving group such as Br, Cl, triflate, mesylate, tosylate, or OC(O)ORs;
Y is a bond, O, S, or NHR4;
Z is a bond; phenylene optionally substituted by 1-4 F, Cl, Br, Cl-C6 alkyl, Cl-C6 alkoxy, Cl-C6 alkyl substituted by 1-13 F, Cl, or Cl-C6 alkyloxy substituted by 1-13 F, Cl, or Br; (C(R4)2)l 2o; or (CF2)l 20; with the proviso that when Z is a bond one of its adjacent Y's is also a bond; and f aryl is a mono- or bi-cyclic aromatic ring cnnt~;n;nS up ~ to 10 carbon atoms and up to 2 heteroatoms selected f rom O, S, and N .

WO 95/28C40 2 1 8 7 7 9 2 P.,l/L~ ~

In the definitions of F2 in Formula Ia, the left-hand bond as depicted attaches to Fl.
Also useful as identifiers are compounds of the Formula Ia ';
F~-(C(E-S~')" Ia' wherein:
F1 i 8 OX or COOH; and the L. ;n;n~ definitions are as in Formula Ia.
Preferred ~ 3q of Formula Ia are those wherein;
10 F~ is FZ is C02H, OH, CH~12, C(O)C~ "(O)X, I~CS, or CX2X:
~2 15 ~ _ ~CH2A-- ~

A--20 ~~ ~ ~
A-- A--OR
~--A or "~A--R

_ and C' is each n~ ly Cl-C20 alkylene or C1-C
alkyl, respectively, unsubstituted or substituted by 1-40 F or Cl, or [O- (CH2)2-3]rJi E is C1-C1~ alkyl substituted by 1-20 F or Cl; Q-aryl where aryl is a bi-cyclic aromatic ring substituted by 1-7 Wo 95n8640 2 ~ 8 7 7 9 2 P~,l/L_ 5 J~' '1 F or Cli or Q-phenyl substituted by 1-5 F, Cl, NO2, or SO2Rs; and Q is a bond, O, -NRsC (O) -, or -OC (O) - .
5 Preferred ~ q of Formula Ia are thoqe wherein -C(E-C'). is represented by -(CH2)~1s-(CF2)1.15F, - (CH2) ~-15- (CCl2) l l5Cl, - (CH2CH2-O) l s-- (CH3CH2CH,O) 1 s-Ar, or - (CH2) 1 l2-O-Ar;
0 wherein Ar is pentafluoro- p.nt~ hl ~ro-, or pental,~ h~nyl, 2,3,5,6-tetrafluoro-4(2,3,4,5,6-p~n~f1llnrophenyl)phenyl~ 2,4,6-trichlorophenyl, 2,4,5-trichlorophonyl, 2,6-dichloro-4-fluorophenyl, or 2,3,5,6-tetrafluorophenyl.
A preferred embodiment of Formula Ia is wherein F1 is COOH, C~N" C(O)C~" 9~O,)CIW" COCI, OH, SH, Cl~,X, or ~l, -.

W0 95l28640 2 1 8 7 7 9 2 wherein F' i5 ~NO 2 ~ N O ~"~, A-- OR' ~A-- ~ --A A-- :
wherein A is -0- or -OC(0)0-;
5 C (E-C' ), is - (CR~ l5- () o l-Ar, - (CR~,) o ls~ (CFl) l-l5-- (CR~2) o-1S- (CF,) l-l5- (CR ,) o ls-H, or - (CH,) l-~o- ( () o-l- (CH2) 1-19) o_~~ (CH,) 0 ,~- () o 1-Ar~
wherein Ar i5 p~ntAflllnro-, p~ntP~hlnro-, or pentabromophenyl, 2,3,5,6- tetrafluoro-4(2,3,4,5,6-pentaf luorophenyl ) phenyl, 2, 4, 6 - trichlorophenyl, 2, 4, 5 -trichlorophenyl, 2, 6 -dichloro-4 -f luorophenyl, or 2, 3, 5, 6-tetraflu~ e~

1~ WO951~0 2 1 8 7 7 9 2 Other preferred . _ iU~ 8 of Forrnula Ia are represented by the forrnulae:
~oz HO~ o~(CH2)0-ls--(CF2)1-~sF
~oz HO~ o--(CH2)1-1s 0-1--Ar o ~0 0 HO ~ oJ~o~ (CH2 )l-ls o-l--Ar HO~ (CH ) --O--Ar O CHNa o--(CHz )O 5--(CFz )I -- sF _ (CH2 )1-15 o-1 Ar If eO~ ~O (CH2)1 1s o-1 Ar O CHI~z O CHN2 -~44 ~
O~le J O~ (CH2 )l l5 Ar 0--(CH2 )1-15 Ar 15 neo~
and SO2CH~2 O (CHz )l -~5 Ar MeO

SO2CH~2 ~ ~
wherein Ar is pentafluoro-, p~nt~fhl-~o-, or pentabL, ,~ yl, 2,3,5,6-tetrafluoro-412,3,4,5,6-p~.n t ~ rophenyl ) phenyl, 2, 4, 6 - t ri chlorophenyl, 2,4,5-trichlorophenyl, 2,6-dichloro-4-fluorophenyl, or 2, 3, 5, 6-tetrafluorophenyl .
Other preferred I aa of Formula Ia are those wherein E-C' is X, OX, or NX2. Such, ,_ ~ are particularly useful for reaction with an E at the end of ~ Wo 95/28640 ~ l 8 7 7 9 2 ~ IIU~
the ,- ' ;n~t~rial synthesis, especially with an E
det~ hl -~ by f luorescence or electron capture, such as dansyl chloride or polyh~ h~n7Qylhalide.
5 Another ~ ' - '; ' of the invention is represented by the following ~
~0--(CH2)-- e~ 1-4 0-2n The compounds of Formula I can be prepared according to 10 the following exemplary schemes or other means known to those skilled in the art.

~ntifi~r Taq Pre~aration X~"~OH lEq,Cs~Co3, DI~F
X~X Br so~c 2 ~o~(c~)n~o~Ar ELECTIIOEEOEl~ E~E~Ob. A~03 X = Cl OR F.

21 877~2 W0 95/28640 r~ s SC}~EME 2 Ifl~r~t; f; ers With PhotolY~ic Cleavaqe Linker~

f q~ OH
COC12 ~BuO_~No2 TOL~E~E
HO~(CH2)n~0~Ar ~ CH2)n~ ,lr Cl O O p~lD~E. CH2CI2 oJ~o~ (CH2 )n~ O, Ar ~Bu0~02 ~;02H. CHaCl2 ~ J~ ,(CHZ~n~ ,~r WOz 2 1 87792 ~
Wo 95128640 SC}IEME 3 T~n~;fiers With Oxidative Cleavaqe IJinker~
~OH
5 MeO~Olle `\~ ~~)n~ O~ ~r PPh3~ DE~D
TOLUE~.~E ~ (CH2~n~0~Ar MeO~OMe o 1. LiOE, TEF/IATE~, O--(CH2~n~o,Ar ~EFLU~ j/q/
2. 50CL2, `11/ \~OMe T0LrE~E REFL~ \ C~2N2 Et2O/CH2Cl2 ~
1~ (CH2)n~o,~r O
.

W0 95/28640 2 1 ~ 7 7 9 2 , ~
--4~--n~; f iF'rS Wi~h ~1 ternative Oxidative Release Linkers ~0~ ~O/(CH2)n~0~Ar ~ (C~2)n\0'Ar P~h3. DE~D. IOL~E~E ~OI~e O O
O--(C ~ ~ ,Ar 1.~[110~. ACEYOI~E/YAYE~ ~/ H2 ~ o Z. O~ALOYL C~LORIDE, Cly\~\ o~
cat. Dll~. CE2~12 ~2~2 Et20/CH2CI2\~
~O~(CI~2)n~ ,Ar O

2 1 87792 " ~
w0 95/28640 ~C~ENE 5 ~ - C ' Taqs F~F 1 BB3, 1~1~ Eo~q~F
F~ 2. CslCo3. 3r(CE2)nC~3 F~O--(C~2)n C~3 0~ F
~CE~13NE 6 lS Identifiers With Photolvtic Cleavaqe T.~nkr-r8 1. COCL2. TOLUE~E
E--C ' TA&
0~~~ ~
PyRL~lNE- CE2Cl2 ~BuO~II--O~ ~O--(CH2)n--CE~g W095/28640 ' 2 1 87792 ~ c~

~CHEME 7 T~9entifier5 With ~h~ tive Cleavaqe Iinkers 01~ E--C' TAG
neo~ TOLUE~E \ ~
--~CH2 )n--CE~3 o one i. LiO~, ThF/~ATER, ~0~--~F
2. sozCl2 Cl~ F~O--(C112)n--CH3 TOLUENE REFLIIX
E~ZO/C~I2CIZ ~ ~0--~CH2)n--CH3 WO 95/28fJ40 2 1 8 7 7 ~ 2 r~ s L ~f ~

The ; ~l~nt; f; .or may comprise one or a plurality o nt;r~l tags. The identifiers will be individual chemical c _ o~n~ (9) which may be distingui3hed one from the other and will unicluely identify different choi~ces and 5 stages. In this manner, very large combinatorial libraries may be prepared with a relatively small number of it~nt;f;ers~ usually fewer than 50 tag6.
During each stage, a, ' in~t;mn of identifiers will be added, which defines the stage and choice. Each 10 ;ti~nt;f;~or Will be bound, either covalently or non-covalently to the bead or to the product, usually the bead . Combinations of identif iers are used to provide a binary or other code at each stage, whereby the choice and stage may be defined. The ;n~t'on of identifiers may 15 include zero or only one; i,~nt; f ier.
~g~
So far as the tags (C-E-_' ) are ronr~rn~d~ the tags which 20 are employed will be rh~rart~rized as follows: by being removable from the bead by means fi~r~nti;ng on Fl or F', preferably by hydrolysis, photolysis or rY;~i~tirn; by being individually differ-~nt;~hl~, usually separable; by being stable under the synthetic conditions; by encoding 25 both stage and choice 80 as to uniquely define the choice of agent used at each stage in the synthesis; desirably, there should be an easy way to identify the various tags with readily-available er~uipment which does not reguire sophisticated technical r~r~hil ities to operate; they 3 0 should be relatively economical and provide a strong signal based on a relatively f ew molecules; and the tags ghould provide suf f icient sensitivity to permit distinguishing the tags from the other ~ _ ~ which may be present during the tag det~rm;n~tinn~
The tags may be structurally related or unrelated, as in a homologous series, repetitive fllnrti~n~l groups, related .
, .,, .. , .. , , . ,,, _ _ _ _ _ _ _ _ W0 9s/286-~0 members of the Periodic Chart, dif ~erent isotopes, ' ;n~tinnq thereof, or the like. The tags may be used as ~l tc of a binary code, so that one tag can define two choices, two tags can define four choices, three tags 5 can define eight choices, five tags can define thirty-two choices, etc. Thus, at each stage of the synthesis, a relatively small number of tags can designate a much larger number of choices. The tags comprising the i~Pnt;f;~rc for each stage may or may not be related to 10 other stages. Each tag for any _ ' ;n~t~ri~l synthesis must allow for being distinguished from all other tags.
In this manner, very large combinatorial libraries may be prepared with a relatively small number of tags, usually fewer than 60 tags, more usually fewer than about 50 tags.
For each bead, there will usually be at least 0 . 01 femtomol, more usually 0.001-50 pmol, of each tag, although lesser or greater amounts may be used in special circumstances. The amount of product may also be at least 20 in the same range and up to at least 10' or more greater, usually being at least 0 . 01 pmol, more usually at least 1. o pmol and generally not more than about 10 nmol .
DPr~nriin~ upon the number of beads, the llumber of stages and the number of choices per stage, the number of 25 products produced will usually exceed 102, more usually 103, and may exceed 101, usually not ~ e~iin~ about loRI
preferably being in the range of about 10~ to 108, more usually 105 to 108.
30 The tags will, for the most part, be organic molecules.
Each tag will usually have fewer than about 100 atoms, more usually fewer than about 80 atoms, generally fewer than about 60 atoms, other than llyd~yc~ rrlll~;n~ a linking moiety which would not be retained on release of 35 the tag from the bead. The linking moiety may be of any size, usually being fewer than about 30 atoms, more usually fewer than 20 atoms, other than l~ydr Uy-:ll. The w09sn8640 21 87792 1~""~ c ~

size of the linking moiety i9 not critical, but one of convenience. The tags may form fAm; 1 i~ of where all of the ~ ~ '~ are of a similar nature or may be -nAtlnn~ of different fAm;l;-'R, where the, ,1~
5 may be Al;rh~t;c~ alicyclic, aromatic, heterocyclic, or , ' inS~t;r-n~ thereof. Distinguishing features may be the number of repetitive units, such as methylene groups in an alkyl moiety, alkyleneoxy groups in a polyalkyleneoxy moiety, halo groups in a polyhalo,~ ~ .u--d, tY- and/or 5-10 substituted ethylenes, where the substituents may involvealkyl groups, oxy, carboxy, amino, halo, or the like;
isotopes; etc.
Taq Analvsis 15 Tags may be removed from the bead using reductive, oxidative, thermolytic, hydrolytic, or photolytic conditions ~rf~nrl;n~ on the nature of the group F2; for example, by ~Y; ~lAt j ~m of a catechol ether with ceric ammonium nitrate or by photolysis of a nitrobenzyl ether 20 or ester or amide, or by other methods, e.g. as shown in Table 1.
Differf~ntiPt;~n of tags can be achieved with physical differences, e.g. ~ clllAr weight of the tags or the 25 cl~ to~raphic rPt~nt;nn time using gas or liquid to~rArhy. Positional isomers may have different r.etPnt;~n times. If positional isomers or steroil 8 are inadequate for physical separation, then one could use varying numbers of substituents, e.g. hAlog~n~ (such as 30 fluorines), methyl groups, oxy groups, or other side chains in conjunction with differing numbers of units, e . g. methylene groups or ethyleneoxy groups, to provide the desired separation. Ratios of radioisotopes could be used, where the radioisotopes provide for differential 35 emission; for example 1~C and 3H. The physical differences, particularly mass number, can provide information about choice and stage.
= . . . . .. .

: . 21 87792 Illstead of l'C/3H ratios, one could use comb;nAtirnq of non-r~l;o~rt;ve ;~otnp~, e.g. -CE~,D", where m i8 0 and up to 3 and n is 3 minus m. For :example, by ~ t~rtinr~ the varying amounts of up to four different methyl groups 5 u~ing mass ~ LLU8~UU~, one could define a large number oE choices.
When E-C' iB H, the tags rht~;n~cl upon release from the support have an active fu~r-tinn~l; ty for reaction with a 10 1 lh~-l ;nrJ reagent whic.h ~ntroduces a ~i~tPrt~hle tag E. Convenieiltly, the fllnrt;rn~1;ty could be a double bond (particularly an activated double bond), hydroxy, thio, amino, carboxy, etc. T~e tag would then be reacted with an exces3 or tne l ~h~l; nrJ reagent to provide 15 the product (E-~H) for analysis. In this way a wide variety of l~h~l;n~ reagents could be used as part of the identifyinr system, which may not be ,~ _ ~; hl ~ with the synthetic strategy for the product of interest. Labeling reagents which may be used for detection include 20 haloaL, -; r~ (e.g., perfluu~ ub~ yl bromide), fluorescers (e.g., dansyl chloride), radioisotopes, chemiluminescers, etc.
While exemplary tags ~nd .~ ct;rn~ have been given, it 25 should be understood ~hat many other combinations could be employed~
3.or~n~1;n~ on the chemical and physical nature of the tags, an appropriate method for separation is chosen, desirably 30 one of various chromatographic pLuceduLe8 ;nrlll~;n~ gas chromatography (GC), liquid u 3,~. tr~raphy (LC) particularly high-performance liquid ull" ~o~raphy (HPLC), thin layer el ~, tt~raphy (TLC), electrophoresis, etc. Instead of a cll~, tr~raphic ~ UC~dULe, mass 35 ~e~:L~, -try may be employed for separation by mass number. Tags include:

w09~n8640 2 1 877q2 for GC: chemically inert organic molecules having the same or different molecular weights ;nrl1lrl;ng alkanes, alkenes, arenes, h:~lor~rh~m~, ethers, alcohols, silanes, thioethers, etc., particularly halogenated compounds, with 5 or without other fllnrtinn~l ities, for electron capture detection or mass spectroscopy detection (MS) with capillary GC separation, and for ~ ' with elements not normally found in organic chemistry (e.g., Sn, Ge) for atom ~mi ~i nn detection with GC capillary seperation;
10 for LC, HPLC or TLC: see above for GC, conveniently linear ethers or hydrocarbons with subætitution by radioisotopes or combin~t;nn~ of radioisotopes for r~; g~8ay detection ~r guitable groups for fluorescence detection a~ter separation;
15 for electrophoresis: see above, particularly f~1n~ tinn=~ otl charged molecules, e.g. cationic or anionic, particularly organic or inorganic acid groups, where the molecule may be further '; f; ~rl by having a detectable r~i;o;~otope or fl~ r- sc~r for detection by 20 electrophoresis;
for mass spectroscopy: see above, particularly different mass numbers due to different isotopes, different numbers of the same functionality or different fllnrt;nn~l;ties, dif f erent members of a b lo~ous series or combinations 25 thereof.
The separation of tags from one another may involve individual terhn;~1~-~ or ' ;n~t;nn~ of tec~iques, e.g.
c~,, tn~raphy and electrophoresis; gas chromatography and 3 0 mass spectroscopy; etc .
The tags of the present invention will have a property which allows ~lPt~oct;nn at very low levels, usuall~ not greater t~an n~n~ 1~, pref erably picomole or less, more5 preferably femtomole or less, in the presence of other which may be present at signif icantly higher levels . For this reason/ ~per; f; ~' atomic substitutions . _ , _ . , . . ,, . , . . ,, _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .

wo 9~/28640 1 ~-/u~ -8 may be used to render the labels eaaily detectable. Such substitutions include:
(a) 6ubstitution by electronegative f~l~ c, e.g.
fluorine or nhlnr;n~ for electron capture detection in 5 conjunction with capillary GC or negative ion mass spectroscopy detection;
(b) substitution by an lln~ element (~n lil~;n r C, H, and 0) for atomic om; RR; nn detection in conjunction with capillary GC;
10 (c) substitution by several, Pl 'R for atomic (~m; RR; nn detection to ~Pt~rm; n~ the ratio between the elements;
(d) substitution by a radioactive element, e.g. 3H, for detection by autoradiography or Rn;nt;ll~t;on counting in 15 conjunction with LC, TLC or electrophoresis;
(e) substitution by a multiplicity of radioactive elements having differing ~m;RR; nnR e.g. 3H and l~C, for detection by autoradiography or 8-;nt;11At;nn rm~nt;n~ to ~ tF~rm;n~
tlle ratio of the di~ferent r~l;o~nt;
For single-element subst; tutinn (a., b., d. above) a separable mixture of A tags whose simple presence or absence can be ~tectad would encode up to 2A different syntheses. For multiple-element substitution (see, c. and 25 e. above) a separable mixture of A tags each having B
disting~l; R~hl~ states (e.g., different 3H/l~C ratios, different Si/Sn ratios) would be able to encode for up to B~' dif f erent syntheses .
30 A wide variety of isotopes exist, where the presence or ratio of isotopes may provide information as to stage and choice. The isotopes may be r~l;n~t;ve or non-rAtl;o~rt;ve. Isotopes of particular interest include deuterium, tritium, 14C~ 32p, l3lI, etc.
By employing mixtures of isotopically-modif ied _ r~R, one can greatly expand the infnrr-t;t~n nht~;nPd from a = = ~

~ wog~640 21 877~2 r~

single tag _ ' which iæ only distinguished by the presence of isotopes. For example, one could prepare a mixture of ratios of hydrogen to deuterium, where the various ratios could differ by as little as 10% each.
By r~rl~c;n~ 11YdLUYe:I1S with another atom, such as flllnrin~, one would then have a varying mixture of lly~uye:lls, deuteriums and fluorines, providing for a large number of different differentiable tags.

Other groups that may be involved could be aromatic rings, which are differentially gubstituted, as to position and functionality. Thus, by having substituted benzene rings, where the position of the substitution and the nature of 15 the substitution can be det~rmin~, one can provide for a plurality of molecules which can be distinguished and can provide for both stage and choice information. For example, if C were constant one could detect and discriminate through the substitution pattern on E when E
20 is a polyhalogenated aromatic ring.
There is also the po~s;h;l;ty to use fluorescent tags.
While f luorescent tags alone may not be suf f icient to define a significant number of gtages with a significant 25 number of choices, as referred to above, by providing for mean8 for separating the fluorescent tagging molecules based on variations in C or C', one can individually detect the tags by their fluorescence.
30 The mixture of tags associated with a particular bead may be detached and 8ubject to an initial separation, where it is desirable to detect each of the tags separately. Once the group of tags has been separated, each of the tags may - then be analyzed based on its particular flln-t;mn~l ;ties 35 and distinctive properties. Various terhn;~u~ which may be used to detect the particular tags include autoradiography or sr;nt;ll~t;on ~mllnt;n~, electron ... . .. , _ . . .. _ .. . , . .. , . .. . . . .. ,, .. .,, . , . , _ . _ _ _ _ _ W0 95128640 2 1 8 7 7 9 2 r~ r~ ~

capture ~-~tert;rn, negative or positive ion mass spectroscopy, infrared spectroscopy, ultraviolet spectroscopy, electron spin resonance spectroscopy, f luorescence, and the like .
Another composition may have at least 6 different markers b~ing associated in a kit or in a common medium, each marker having a disting--i Al~hl t moiety which is subst~nt;~l1y chemically inert, differing from each other 10 in molecular weight, said markers of the formula:
~1) A-{~- (T)tt or ~T)"-Q or ~1 (T)tt-~2}
where 1~ is a linking group which has a functionality for 15 bonding to a solid gupport and a functionality for tlt~t~` t from the solid support, which may be included in the functionality for~bonding to the solid supporti ~ is a distinguishing group, which allows for the 20 distinguishing by a physical characteristic of each of the markers from each of the other markers by other than fluorescence, 80 as to provide a set of markers which allows for coding a multistep synthetic ~Lv~eduLe, and includes any L~ ;n;n~ fllnrt;rn~l group after tlt-tarl -, 25 which was previously actqOr; ~tt~t~ with the linked group;
~1 and 1~2 are portions of the distinguishing group, which together define a distinguishing group and when joined together come within the definition of Q;
T is a ~ t~-ct~hl t~ group, which when attached to the distinguishing group allows for the low level detection of the marker, where the detectable group may be present on the marker in the kit or may be added later to the 35 distinguishing group and if attached to the linking group, includes any L~ ;n;nJ functional group after det~t~ t, ~hich was previously associated with the linking group;

~ W095/28640 2 1 ~7792 ~ 5-l ~

and a is 0 or 1, indi~ating that the det-~rt~hl~ group may or may not be present;
(2) ss-(A'-{~-(T) or (T)o~~ or ~-(T)-~2})~
where all of the symbols have been defined previously, except as follows: S8 is a solid support;Q' is a linking group covalently bonded to 88; and ,~ is an integer as to each solid support and is at least six and usually not 10 more than about 30;
(3) A"-{~"-(T)o or (T~)o~~ or ~'l-(T)o~~2}
where all of the symbols have been defined previously, 15 except as follows: ~" is 1~YdL~YC:~ or the residue of the linking group after photolytic cleavage, ~l;m;n:~t;~n or other rh~m; f~l reaction which regultg in ~l~t~ ` t from the solid support; ~" or ~1 is ~ or ~1 or a modified 1~ or ~1, respectively, as a result of the det~` o~ the 20 marker from the solid support; T" is T or a modified T as a result o~ the tl~t~` ' of the marker from the ~olid support;

(4) To~A'Il~{~ll~(T)o or (T~)o~Q or L'l-(T)o~~2}
where all of the symbols have been defined previously andA' " is a bond or the ~. ;n;n~ portion of the linker group after ;:ltt~` t to T; with the additional limitation that only one a is 1.
Assays To determine the characteristic of interest of the product, a wide variety of assays and te~ hn;q~ may be employed .
Fre5~uently, in screening the beads, one will use either single beads or mixtures of beads and ~l~t~rm; n~ whether W095128640 2 1 8 7 7 9 2 . ~

the bead or mixtures show activity. Thus, the mixtures may involve 10, 100, 1000 or more beads. In this way, large groups of ~ _ 'q may be rapidly screened and segregated into smaller groups of compounds.
One technis~ue is where one is interested in binding to a particular biomolecule such as a receptor. The receptor may be a single molecule, a molecule associated with a microsome or cell, or the like. Where agonist activity is 10 of interest, one may wish to use an intact organism or cell, where the response to the binding of the subject product may be measured. In some instances, it may be desirable to detach the product from the bead, particularly where physiological activity by transduction 15 of a signal is of interest. Various devices are available for detecting cellular response, such as ~ a microphysiometer, available from Molecular Devices, Redwood City, CA. Where binding is of interest, one may use a labeled receptor, where the label i8 a fluorescer, 20 enzyme, radioisotope, or the like, where one can detect the binding of the receptor to the ~ ~_ ' on the bead.
~lternatively, one may provide for an antibody to the receptor, where the antibody is labeled, which may allow for amplification of the signal and avoid rh~n~;n~ the 25 receptor of interest, which might affect its binding to the product of interest . Binding may also be ~''t~'rln; n/~
by disp~ ~l of a ligand bound to the receptor, where the ligand is labeled with a detectable label.
30 In some instances, one may be able to carry out a two-stage screen, whereby one first uses binding as an initial screen, followed by biological activity with a viable cell in a second screen. By employing L~- ' ;n:~nt technislues, one can greatly vary the genetic r~r~h;l;ty Of 35 cells. 3ne can then produce o~ g~n~llq genes or e, ~y~l. us transcriptional regulatory sequences, so that binding to a surf ace me~nbrane protein will result in an observable W095128640 21 877~2 ,~." c~

signal, e.g. an intrar~ lAr signal. For example, one may introduce a leuco dye into the cell, where an enzyme which transforms the leuco dye to a colored product, particularly a fluorescent product, becomes expressed upon

5 appropriate binding to a surface membrane, e.g.
Al Art( IRi~lA Re and di~ l A~ tr~Ridylf luorescein . In this manner, by associatiny a particular cell or cells with a particular particle, the fluorescent nature of the cell may be ~ t~rm;n~d using a ~.CS, so that particles carrying 10 active ~ may b~ ic~ntified. Various techni~aues may be employed to ensure ~hat the particle remains bound to the cell, even where th~ product is released f rom the particle. For example, one may use Ant;ho~ R on the particle to a surfa~e membrane protein, one may link 15 avidin to the surface of the cell and have biotin present on the particle, etc.
Assays may be performed stagewise using individual particles or groups of particles or combinations thereof.
20 For example, after carrying out the combinatorial syntheses, groups of about ~o to 10, 000 particles may be segregated in separate vessels. In each vessel, as to each particle a portion ~f the product bound to the particle is released. lhe rArtionA1 release may be as a 25 result of differential linking of the product to the particle or using a lim~tad amount of a reagent, condition or the like, so that the average number of product molecules released per particle is less than the total number of product molecules per particle. One would then 30 have a mixture of products in a small volume. ~he mixture could then be used in an assay for binding, where the binding event could be inhibition of a known binding ligand binding to a receptor, activation or inhibition of a metabolic process of a cell, or the like. Various assay 35 conditions may be used for the detection of binding activity as will be described sllhRPsrl~ontly~ Once a group is shown to be active, the individual particles may then wo 95n8640 be screened, by the same or a different assay. One could of course, have a three- or four-stage procedure, where large groups are divided up into smaller groups, etc. and finally single particles are rscreened. In each case, 5 portions of the products on the particles would be r~leased and the resulting mixture used in an appropriate assay. The assays could be the same or dif~erent, the more sophisticated and time consuming assays being used in the later or last stage.
One may also provide for s~atial arrays, where the particles may be distributed over a h~ y~ ` plate, with each well in the hul~, ` having O or l particle.
15 The subject methodology may be used to find chemicals with catalytic properties, such as hydrolytic actiYity, e.g.
esterase activity. For this purpose one might embed beads in a semisolid matrix ~ ed by dif fusible test substrates. If the catalytic activity can be detected 20 locally by processes that do not di6turb the matrix, for example, by changes in the absorption of light or by ~ter~;rrl of fluorescence due to a cleaved substrate, the beads in the zone of catalytic activity can be isolated and their labels decoded.
Instead of catalytic activity,, __ ~q with inhibitory or activating activity can be developed. t`~ ~lq may be sought that inhibit or activate an enzyme or block a binding reaction. To detect beads that inhibit an enzyme, 30 which beads have an attached product with this desirable ~L-,~e, Ly, it is advantageous to be able to release the products from the beads, ~n;~hl ;n~ them to diffuse into a s~m;qnl;f~ matrix or onto a filter where this inhibition, activation or blork; nJ can be observed. The beads that 35 :Eorm a v; S:nl~l; 7~ or otherwise detectable zone of inhibition, activation or hlork;ng can then be picked and the tags decoded. In this case it is nProqs~ry that a wossns640 21 87792 .~ 7 portion of the synth~R; 7~d products be attached to the beads by cleavable linkages, preferably a photolabile linkage, while a portion of the tags remain attached to the bead, releasable after picking by a different means 5 than before.
A dialysis membrane may be employed where a layer o~ beads is separated from a layer of radiolabeled ligand/receptor pair The bead layer could be ;rrAr7.;;7t~7 with ultraviolet 10 light and the product released from the bead would diffuse to the pair layer, where the radiolabeled ligand would be released in proportion to the affinity of the compound for the receptor. The rA~7.;nlAhPl~d ligand would diffuse back to the layer of beads. Since the radiolabel would be 15 proximal to the bead, beads asgociated with radio~m;Rginn would be analyzed.
Of particular interest is f inding products that have biological activity. In some Arpl ;--At;nnR it is desirable 20 to find a product that has an effect on living cells, such as inhibition of microbial growth, inhibition of viral growth, inhibition of gene expression or activation of gene expression. Screening of the ~~ ,.,u-~ds on the beads can be readily achieved, for example, by ~ 7~7 7 n~ the 25 beads in a semisolid medium and the library of product molecules released from the beads (while the beads are retained~ ~n~7hl;n~ the ~ 7R to diffuse into the surrounding medium. The effects, such as plaques within a bacterial lawn, can be observed. Zones of growth 30 inhibition or growth activation or effects on gene expression can then be v; R17Al; 7~'~7 and the beads at the center of the zone picked and analyzed.
One assay scheme will involve gels where the molecule or 35 system, e.g. cell, to be acted upon may be em.bedded substAnt;Ally ~ , eouRly in the gel. Various gelling agents may be used such as polyacrylamide, agarose, 21 ~7792 Wo 95/286-~0 gelatin, etc. The particles may then be spread over the gel so as to have suf f icie~t separation between the particles to allow for individual ~f-tect;rn. If the desired product is to have hydrolytic activity, a 5 substrate is present in the gel which would provide a fluorescent product. One would then screen the gel for fluorescence and Ir-^h~n;c~lly select the particles ~q80r;~ with the fluorescent signal.
10 One could have cells f:~mh~ d in the gel, in effect creating a cellular lawn . The particles would: be spread out as indicated above. Of course, one could place a grid over the gel ~l~f; n; nrJ areas of one or no particle . If cytotoxicity were the criterion, one could release the 1~ product, incubate for a sufficient time, followed by spreading a vital dye over the gel. Those cells which absorbed the dye or did not absorb the dye could then be distinguished .
20 As ;n~;r~ted above, cells can be genetically ~nrJin.-~red 80 as to indicate when a signal has been transduced. There are many receptors f or which the genes are known whose expression is activated. By inserting an t:~UUt:lIUU8 gene into a site where the gene is under the transcriptional 25 control of the promoter responsive to such receptor, an enzyme can be produced which provides a detectable signal, e.g. a fluorescent signal. The particle associated with the fluuL~:sc~ L cell (8) may then be analyzed for its reaction history.
T.; h~;l~ieg and Kits For convenience, libraries and/or kits may be provided.
The libraries would comprise the particles to which a library of products and tags have been added 80 as to 35 allow for screening of the products bound to the bead or the libraries would comprise the products removed from the bead and grouped singly or in a set of 10 to 100 to wo 95t28640 1000 members for screening. The kits would provide various reagents f or use as tags in carrying out the library syntheses. The kits will usually have at least 4, usually 5, different compounds in separate rnnt~;nprsr 5 more usually at least 10, and may comprise at least 102 different separated organic . '~, usually not more than about 102, more usually not more than about 36 different ~ '~. For binary determ;n~t;oncl the mode of detection will usually be common to the mln~c 10 associated with the analysis, so that there may be a common ~ hnrel a common atom for detection, etc.
Where each of the i~ nt;flers is pre-prepared, each will be characterized by having a distinguishable composition f.nrn~;nr choice and gtage which can be determined by a 15 physical mea2.u~ and ;nr~ in~ groups or all of the ~lc 8haring at leagt one common functionality.
Alternatively, the kit can provide r~rt~ntR which can be c ' ;n~d to provide the various i~ nt;f;ers. In this 20 situation, the kit will comprise a plurality of separated first functional, frequently bifllnrt;nn;-l, organic 1R~ U5ually four or more, generally one for each stage of the synthesis, where the fllnrt;nn~l organic compounds share the same functionality and are 25 distingll;Rh~hle as to at least one determinable characteristic. In addition, one would have at least one, usually at least two, second organic compounds capable of reacting with a functionality of the fllnrt;nn;ll organic Jullds and capable of forming mixtures which are 30 distinguishable as to the amount of each of said second organic ~ , IR. For example, one could have a glycol, amino acid, or a glycolic acid, where the various bifllnrt;nn~ aR are distinguished by the number of fluorine or chlorine atoms present, to define stage, and 35 have an i~ ' -h~n~, where one jn~ ' h~n~ has no radioisotope, another has I'C and another has one or more 3H. By using two or more of the ic ' th~n~R, one could wo 95/28640 2 1 8 7 7 9 2 . ~ I/L.~

provide a variety of mixtures which could be ~iptprminpd by their radioemissions. Alternatively, one could have a plurality of second organic, ~olln~lC, which could be used in a binary code.
A8 indicated previously one could react the tags after release with a molecule which allows for detection. In this way the tags could be quite simple, having the same functionality for li~king to the particle as to the lO detectable moiety. For example, by being linked to a hydroxycarboxyl group, a hydroxyl group would be released, which could then be estPrif;Pd or etherified with the molecule which allows for tlptectinn~ For example, by u6ing comb;n~ti~n~ of fluoro- and chloroalkyl groups, in 15 the binary mode, the number of fluoro and/or chloro groups could determine choice, while the number of carbon atoms would indicate stage.
Groups of compounds of particular interest include 20 distinguishing groups such as C joined to a substituted ortho-nitrobenzyloxy group, indanyloxy or fluorenyloxy group, or other group which allows for photolytic or other selective cleavage. The distinguishing group may be an alkylene group of from 2 to 20 carbon atoms, 25 polyalkyleneoxy, particularly alkyleneoxy of from 2 to 3 carbon atoms, cycloalkyl group of from 4 to 3 carbon atoms, haloalkyl group, particularly fluoroalkyl of from 2 to 20 carbon atoms, one or more aromatic rings and the like, where the distinguishing group provides for the 30 discr;m;n~tinn between the various groups, by having different numbers of units and/or substituents.
Individual particles or a plurality of particles could be provided as articles of commerce, particularly where the 35 particle (s) have shown a characteristic of interest.
}3ased on the associated tags, the reaction history may be decoded. The product may then be produced in a large 2l 87792 W095/28640 I~"L~_r~

synthesis. Where the reaction hiatory unequivocally def ines the structure, the same or analogous reaction series may be used to produce the product in a large batch. Where the reaction history does not unambiguously 5 define the structure, one would repeat the reaction history in a large batch and use the resulting product for structural analysis. In some instances it may be found that the reaction series of the ncmhin~tr1~ial chemistry may not be the pref erred way to produce the product in 10 large amounts.
An ~ ' '; of this invention~is a kit comprising a plurality of separated organic ~ iullds, each of the ~R characterized by having a disting l; qh~hle 15 composition, .onrnr~in~ at least one bit of different information which can be det~rm;n~l by a physical mea~iu~ ~ , and sharing at least one common flln~t;nn~1;ty A preferred embodiment is a kit comprising at least 4 different functional organic _ '~.
More preferred is a kit wherein said functional organic compounds are of the f ormula:
F1-F2_C_E_~ I
and the f ormula:
F1-F2-(C(E-C')a)b (Ia) where F1-F2 is a linker which allows for at~` to and ~'t~l' t from a solid particle; and C-E-C' is a tag member which can be det~rm;nf~d by a physical mea~ul t, ~per;~lly wherein said fllnrt;nn:ll 30 organic _u--ds differ by the number of methylene groups and/or h~lo~nc, nitrogens or sulfurs present.
Also preferred is a kit wherein the C-E-C' portion is removed photoqh~m;f~lly or a kit wherein the C-E-C' 35 portion is removed oxidatively, hydrolytically, thermolytically, or reductively.

wot~640 2 1 8 7 7 9 2 , ~"~,~ r, ~ t In one : tl; the invention is a composition comprising at least 6 dif~erent ,- _ ^ntg, each component having a disting~l; Ah~h~ e moiety. The components may be characterized by each moiety being substAnt;~lly 5 chemically stable or inert and having an it~t~nt; f; ~hle characteristic different from each o~ the other -;et;t~A.
Each moiety is joined t:o a linking group having an active fllnt~t;t~n:~l;ty capable of forming a covalent bond through a linking group to individL~Llly separable solid surfaces, 10 or joined to a group whi~ is detectable at less than 1 -le, with a proviso t~lat when the moieties are joined to the linking group, ~h~ _ _ ^ntA are physically segregated. Preferably, the solid supports are beads.
In one t~mh~t~; ' each component comprises molecules of 15 d~fferent ., 'A bound to individual separable solid surfaces, wherein the ~ t~lllt~A on the solid surfaces.
Preferably, the moieties of the invention define an 1- 1 o~ol-A 8erieg and/or a geries of substitutions on a core molecule.
The invention herein is also directed to a uu--d library comprising at least one hundred uni~ue solid supports. In this ~ ' l ibrary each solid support has (1) an individual com~oun~ ~ound to the solid support as 25 a major ~ , t~ boun~ to tke support; and (2) a plurality of tags e . g . tags incapa~l~ of being sequenced, where the tags are individual tag molecules which are physically distingll; Ah~hle in being physically separable and are substituted 80 as to be detectable at less than about a 30 nanomole or have a ~unctional group for bonding to a substituent which is detectable at less than about at nzlnt 1 t~ . Preferably, in the, _ t~ library each solid support has at least about 6 tags. In another . ' _ '; , in the , _ ' library the tags def ine a blnary code 35 t~nt~ot~;n~ the synthetic protocol uged for the gyntht~A;~;n~
of the: , ' on the solid support.

w09i~640 2 1 ~ 7 7 9 2 This invention also provides a method for rlP~P~n;n;nrJ a synthetic protocol encoded by separable physically different tags in a series and defining a binary code. In this method at least two tags are employed to def ine each 5 stage of the synthetic protocol, there being at least six tags. The step of the method comprises separating tags by means of their physical differences and detecting the tags . The synthetic protocol is def ined a binary code of dif ferent tags .

of this invention may be useful as analgesics and/or for the treatment of ;n~li t~ry disease, P~per;iql ly in the case of the azatricyclics acting as iqnt~rJ~n; AtS of the neurokinin l/bradykinin receptor.
15 Members of the bPn7P~;i37opine library may be useful as a muscle relaxant and/or tranquilizer and/or as a sedative.
Members of the 23.5 Million Mixed Amide Library (Example

6) may be of use in the treatment of hypertension or Raynaud~s ~ylldr. by acting as endothelin iqnt~rJ~n;~ts.

PEPTIDE LIBRARY
In order to encode up to 109 different syntheses, one could prepare 30 different i~lPn~;fiers which carry individual 25 tags capable of being separated one from another by capillary GC. For Pnro~;nrJ a smaller number of syntheses, fewer ;tlPn~;~;Prs would be used. The tags would normally be ~ e~ d from commercially-available chemicals as evidenced by the following illustration.
30 ~-~ydroxyalkenes-1, where the number of methylene groups would vary from 1 to 5, would be reacted with an iodoperf luoroalkane, where the number of CF2 groups would be 3, 4, 6, 8, 10, and 12. By employing a free-radical catalyst, the iodoperfluuLu~ic,,Lu-- would add to the double 35 bond, where the iodo group could then be reduced with llydru~ ll and a catalyst or a tin hydride. In this manner, 30 different tags could be prepared. The chemical .

wo95/286~0 2 1 87 procedure i8 described by ~q7el~;n~ and Steele, J. Chem.
Soc. (Jondon), 1199 ~1953); srace, J. Fluor. Chem., 20, 313 (1982). The highly fluorinated tags can be easily ~l~t~oct~ by electron capture, have different GC ret~nt;~n 5 times, 80 that they are readily separated by r~ ry GC, are chemically inert due to their fluorinated, hydrocarbon structure and each bears a single hydroxyl functional group for direct or indirect att~ to particles.
10 sefore attachment to _ ~ precursors, the tags (referred to as T1-T30) would be activated in a way which i8 appropriate for the chemical inteL~ tPq to be used in the ;n~t~rial synthesis. sy appropriate it is ;nt~-nrlod that a fl~nrtinn~l;ty would be added which allows 15 for ready att~:' by a chemical bond to a precursor or to the bead matrix itself. The activation process would be applied to each of the 30 different tags and allow these tags to be chemically bound, either directly or indirectly, to intermediates in the 20 co~binatorial uu-ld synthesis. For example, a carboxy derivative could be used for rollpl;ng and upon activation the resulting carboxy group would bond to the particle.
In the case of a, ' ;n~tr~rial synthesis of a peptidic 25 l~ _ ~ui~d or other structure made of amide-linked organic LL _ q, the encoding process could consist of addition of a carboxylic acid-eciuipped linker. For example, the tag would be coupled to the ~.-butyl ester of Q-nitro-~-uclLllu~yiJell~yl bromide in the ~,~selice of sodium hydride.
30 Tlle ester would then be hydrolyzed in dilute trifluoroacetic acid.
~ctivated ;~Pnt;f;~rs would be coupled to intermediates at each stage in the ~ ' ;n~t~)rial ~ _ol_n~l synthesis. The 35 ortho-nitrobenzyl ether part of the activated identif iers i~ used to allow photochemical detq~l : of the tags after ~et;n~ the combinatorial synthesis and selecting 1--w09sl28640 ~ 87792 the most desirable ~ ul.ds. The detached tags would then be decoded using capillary GC with electron capture detection to yield a history of the synthetic stages used to prepare the , r7 selected.

7While there is an almost unlimited set of chemical stages and methods which could be used to prepare combinatorial libraries of , ~ Jul-ds, we will use rol7rl i n~ of Q-amino acids to make a ` ;ns7tr,rial library of peptides as an 10 example of an ~rpl;r~t;c~n of the PnrQr7;n~ thr,r7~7lo~y. In this example, we will describe preparation of a library of pPnt-7rPrtide8 having all combinations of 16 different amino acids at each of the f ive residue positions . Such a library would contain 16~ men7bers. To uniquely encode all me7l7bers of this library, 20 dPt~r7~-7hl e tags (Tl-T20) as described above would be required.
To prepare the encoded library, we would begin with a large number (~106) of polymer beads of the type used for 20 Merrifield solid phase synthesis and funct; ~r,ns7l; ~ed by free amino groups. We would divide the beads into 16 equal portions and place a portion in each of 16 different reaction vessels (one vessel for each different cY-amino acid to be added). We would then add a small portion 25 (e.g., 1 mol~) of 7r7Pnt;f;Pr~ to each of the amino acid derivatives (e.g., ~moc amino acids) to be coupled in the first stage of the, ~ ;nz7torial synthesis. The specific c ' ;nAt;on of the tags incorporated into the ;~7Pnt;f;ers added would represent a simple binary code which 30 ;r7Pnt;f;PR the amino acid used in the first stage of synthesis. The 16 amino acids added would be indicated by numbers 1-16 and any such nun7ber could be represented chemically by ,- :n;7t;rnc of the first four tags (Tl-T4) .
In tables 2 and 3, a typical Pnrorl7; n~ scheme is shown in5 which the presence or absence of a tag is indicated by a or a 0, respec ~ ively . The letter T may represent Wo 95128640 2 1 8 7 7 q 2 either the the tag or the i~nt;f;er incorporating that tag .
5 Tabl~ 2. A typical l~n~o~;n~ scheme.

Amino Acid added in fir~t T4 T3 T2 Tl ~t~ge 10 Number l (e.g., glycine ~ 0 0 0 0 Num,ber 2 (e.g., alanine) 0 0 0 Nu,mber 3 (e.g., valine) 0 0 l 0 Number 4 (e.g., serine) 0 0 Nu,mber 5 ( e . g., threonine) 0 l , 0 0 .

Nu,mber.16 (e.g., tryptophan) 20 We would then carry out a standard dicyclohexyl-r~rho~;;m;r~ (DCC) peptide Collrl;n~ in each of the 16 vessels using the Fmoc amino acids admixed with small amounts of the Pnt~or8 ~ n~ activated; ~l~n t ; f i ers as; n~ t ~d above. During the couplings, the amino acids as well as 25 small amounts (e.g., 196) of the i~i~nt;f;~s would become chemically bound to inte~ t~ attached to the beads.

Next the beads would be thoroughly mixed and again separated into 16 portions. Each portion would again be 30 placed in a difierent reaction vessel. A second amino acid admixed with appropriate new activated identifiers (T5-T8) would be added to each vegsel and DCC rollrl ;n~
would be carried out as before. The particular mixture of wo 9s~640 2 1 8 7 7 9 2 1~l" -~, the incv, ~vLc.ted tags (T5-T8) would again represent a simple binary code for the amino acid added in this, the second stage of the ~ ' ;n~ rial synthesis.

5 T~le 3. A typical ~n~orl~n~ scheme.

Amino Acid ~dded in ~3econd T8 T7 T6 T5 st~ge Number l ( e . g ., glycine 0 0 0 0 10Number 2 (e.g., alanine) 0 0 0 Number 3 ( e. g., valine) 0 0 1 o Number 4 ( e . g., serine) 0 0 Number 5 ( e . g ., threonine ) 0 1 0 0 Number 16 (e.g., tryptophan) After the 16 courl;n~a of stage 2 are, ~lete, the beads would be again mixed and then divided into 16 new portions for the third stage of the gynthesis. For the third stage, T9-T12 would be used to encode the third amino acid bound to the beads using the same scheme used for stages 1 and 2. After the third collrl ;n~s, the procedure would be repeated two more times using the f ourth amino acids with T13-T16 and the fifth amino acids with T17-T20 to give the entire library of 1,048,576 different peptides bound to beads.

Although the above beads would be visually 30 indistinguishable, any bead may be chosen (e.g., by 8~1ect;n~ baged on the interesting chemical or biological _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ W0 9~/28640 properties of its bound peptide or other target molecule) and its synthetic history may be learned by detaching and ~C~co~1 n~ the AFlAoCi ;It~fl tags .

5 The precise method used to detach tags will depend upon the particular linker used to rh~m;r:~lly bind it to inte~ ;Ates in the combinatorial synthesis of the target compound. In the example above, the ortho-nitrobenzyl carbonate l; nk;lg~A, which are known to be unstable to ~300 nm light (ohtsuka, et al., .J. Am. Chem. Soc., 100, 8210 [1978] ), would be cleaved by photochemical irradiation -of the beads. The tags would then diffuse from the beads into free solution which would be injected ~ to a capillary gas chromatograph (GC) eqntrped with a 15 sensitive electron capture detector. Since the order in which the tags (T1-T20) emerged from the GC and their retention times under standard conditions were previously determined, the presence or absence of any of T1-T20 would be directly determined by the presence or absence of their 20 peaks in the GC c:1l. to~ram. If 1 and O represent the presence and absence regpectively of peakg corrP~pnr~; nr~
to T1-T20, then the ~:11LI- to~ram can be taken as a 20-digit binary number which can uniquely represent each possible synthesis leading to each member of the peptide 25 library. The use of 1~A1 nrArhnn tags which are safe, economical and detectable at 8llhri 1 e levels by electron capture detection makes this capillary GC method a particularly convenient onrn~;n~ scheme for the purpose.

Wo 9s/28640 2 1 8 7 7 9 2 A ~ 1 1 ~ C I
As an example of using the PnrOr-i;n~ scheme for the pl~nt~rPr~i~P library above, a particular bead is irradiated with light to detach the tags, the solubilized labels injected into a capillary GC and the following 5 chromatogram obtained ( "Peak" line):

L-bel Z3 19 1t 17 16 15 14 13 12 11 10 9 a 7 6 S 4 3 2 1 GC Inject 10 Peflk ~in~ry 1 1 1 1 0 1 0 3 C G ~ 1 0 0 0 1 0 0 1 0 st-~e 5 4 -~ 2 15M Tryptophfln Threonin~ S~rin~ Alflnine V~Line The "~abeln line diagramg the GC ~l~rl ~ o~ram where T20-T1 20 peaks ( ¦ ) are to be found (note the injection is given on the right and the chromatogram reads from right to left).
The "Peak" line represents the presence of labels ~T20-T1) as peaks in the c~-- ~s~ram. The "Binary" line gives presence (1) or absence (0) of peaks as a binary number.
25 The "Stage" line breaks up .he binary number into the five different parts Pn~ ths five different stages in the synthesis. Finally, thf~ I~j~n line gives the identity of the amino acid which was added in each stage and was given by the binary code in the "Binary" line above.

W0 95/28640 2 l 8 7 7 9 2 , ~", EXaN rLE 2 In the next illustration, the tags employed are ~hyletherg of linear alkyl~ -diols. The diol 5 would have ~ + 2 carbon atoms, where N designates the stage. The methyl group would be a radiolabeled reagent which would have any of a variety of 3}~ C ratios from 1/l to m/1, where m is the number o choices. The double rA~;n1~h-1 allows for accurate ~uantitation of the tritium 10 present in the tag. By having 10 different alkylene groups and lO different rA~l;na~At;ve label ratios, 101~
uni~ue ten-member sets of tags are generated. Tags would be attached by first reacting them with activating agents, e.g. rhnsj_n~ to form a chloroformate, followed by 15 reaction with the Fl-F2 ~ Ant. In this case, Fl-F2 is the o-nitro-p-carboxy-benzyl alcohol protected as the t-butyl ester. Each time a synthetic stage is carried out, the de-esterified ;tlAnt;f;~r is added directly to the bead, which _as covalently bonded amine or hydroxyl 20 groups, to form amides or esters with the~acid activated using standard chemistry, e.g., rAArhnrl;;m;rl~ coupling r ~hnrlnlogy At the end of the se~uential synthesis, the beads are then screened with a variety of receptors or enzymes to determine a particular characteristic. The 25 beads demonstrating the characteristic may then be ; Anl A t d, the tags ~ ~ A rh d and separated by }I~LC to give a series of glycol monomethyl ethers which may then be analyzed for r~rl;na~At;vity by standard radioisotope Wo 95/2O640 2 1 8 7 7 9 2 F ~ ~

nt;f;cation methodg. For example, if the first and second tags to elute from the HPLC column had 3EI/l'C ratios of 5.1 and 7:1 respectively, then the product which showed activity had been synthesized by reagent number 5 in 5 stage 1 and reagent number 7 in stage 2.

E~al5PLE 3 2401 Pe~tide Librarv The jrlpnt;f;~rs employed were 2-nitro-4-carboxybenzyl, 10 O-aryl substituted ~o-hydroxyalkyl carbonate, where alkyl was of f rom three to 12 carbon atoma and aryl was (A) p~nt~chlr~rophenyl, (B) 2,4,6-trichlorophenyl, or (C) 2,6-dichloro-4-fluorophenyl. The tags are designated as NAr, wherein N is the number of methylene groups minus two and 15 Ar is the aryl group. Thus, tag 2A has a butylene group bonded to the p-~nt~hl..L~yht:llyl through oxygen. The subject tags can be easily flF~tert~ using electron capture gas chromatography at about 100 fmol.

20 In the subject analysis, the tagging ~ c~ are arranged in their GC elution order. Thus the tag which is retained the longest on the GC column is designated Tl and is ~or;~to~l with the least significant bit in the binary synthesis code number, the next longest retained tag is 25 called T2 reprP~nt; n~ the next least signif icant binary bit, and BO on. Using an 0.2mM x 20M methylsilicone ~-~rill~ry GC column, eighteen well-resolved tags were .ht:~;n~ where Tl through T18 corresponded to lOA, 9A, 8A, . _ . ... _ _ . . , .. . .. . . . , . . .. . _ _ ... .... _ .. . . . ... _ _ _ _ W0 9s/28640 2 t 8 7 7 9 2 7A, 6A, 5A, 4A, 3A, 6B, 2A, 5B, lA, 4B, 3B, 2B, lB, 2C, and IC, respectively.
An encoded combinatorial library of 2401 peptides was prepared. This library had the amino acid sequence N-RRnI~ -beadl where the variable X residues were D, E, I, R, L, Q, or S (single letter code). The 4 glycines served as a spacer between the encoded amino acid sequence and the bead. The ~ ' in~torial library included the sequence HiN-RT,T.~RRn~, part of ~ the 10 amino acid epitope which is known to be bound by 9E10, a monoclonal antibody directed against the human C-myc gene product. For ~.nrnr~ins this library, three binary bits were sufficient to represent the seven alternative reagents for each stage. The code was as follows: 001 = S; 010 = I; 011 =
K; 100 = L; 101 ~ Q; 110 ~ E; 111 = D.
The lii:rary was synth(~; 7~cl by first preparing the constant segment of the library H~RRnTlGGt~ -bead on 1 5 g of 50-90~ polystyrene synthesis beads functinn~l i 7e~1 with 1.1 me~I/g of zlm;- thyl groupg uging gtandard solid phase methods based on t.-butyl side-chain protection and Fmoc main chain protection ~Stewart and Young, "Solid Phase Peptide Synthesis", 2nd edition, Pierce Chemical Co., 1984). After deprotecting the Fmoc groups with diethylamine, the beads were divided i~to seven 200 mg fractions and each fraction placed in a different Merrifield synthesis vessel mounted on a single wrist-WOgS/28640 21 87792 r ~

action shaker. The beads in the seven vessels were processed ;ntl~r~n~.ntly as follows (see Table 3-1). The letter T in this example refers to the tag or to the nt;f;er incorporating that tag.

TABI.E 3-1 Ves Step 1 Step 2 Step 3 Step 4 8~1 10 No.

196Tl DIC, wash Fmoc(tBu~S, Anh. Wash 2l~T2 " FmocI, Anh. "

3l~Tl, T2 " Fmoc ~Boc) K, Anh . '~

4l~T3 " Fmoc~, Anh. "
l~Tl,T3 " Fmoc(trityl)Q~ "
Anh .

61~6T2,T3 n Fmoc(t-butyl~E, "
Anh .

719~Tl,T2,T3 " Fmoc(tBu)D, Anh. "

In accordance with the above procedure a sufficient amount 20 of the i~l~nt;f;prs listed in step 1 were attached via their carboxylic acids using diisopropylcarbodiimide to tag about 19~ of the free amino groups on each bead in the corr~pnn~;n~ vessel. The L. ;n;nS free amino groups on each bead were then coupled in step 3 to N-protected amino 25 acid anhydrides. After washing with methylene chloride, isu~u~a.lol, and N,N-dimethylformamide, the beads from the 6even vessels were ~ ` ;n~-~l and thoroughly mixed. At this point the library ha~ seven members.

W095l28640 21 8 7 792 r~ s L ^~ ~
--~o--After Fmoc deprotection (diethylamine), the beads were again divided i~to seven vessels and processed as before except that in place of the ;ll~ntif;Prc used previously, ;~Pnt; f;~r8 reprPqPnt;n~ the gecond stage (T4-6) were 5 used. By rPrPAtin~ the procedure two more times, using ~rqpnt;f;prg T7-9 and then T10-12 AnAlo~ollcly~ the entire uniquely encoded library of 74=2401 different peptides was prepared using only 12 i~pnt; f; ers .

10 To read the synthesis code from a single selected bead, the bead was first washed four times in a small centrifuge tube with 100 ~LL portions of DMF, and then rPRllqpPn~lP~ in IlL of DMF in a Pyrex capillary tube . Af ter 2 hrs of photolysis with a Rayonet 350 ~m light source, the tags 15 released from the bound ;~1Pnt;f;~rs were silylated using about 0 .1 ~L bis-trimethylsily] ~'Ptr ~P and the solution inj ected iIlto a Hewlett Packard , Ar; 11 Ary gas chromatograph eç[~;rpPcl with an 0.21nM x 20M methylqil;c-nP
fused silica capillary column and an electron capture 20 ~letector. The bi~ary synthesis code of the selected bead ~was direc~ly determined from the cll~ to~ram of the tags which resulted.

wo 95,28640 2 1 8 7 7 9 2 1 ., .,~ s EXANP~ 4 R!~n7~ ; A7f~r~; ni~ Librarv A ,~ ' in~t~ rial b~n7~ 7~pjn~ library comprising 30 of the formula VIII
Cl HO~ \N--Rl VIII
, 10 R
wherein:
R is CH3, CH(CH3)" CH2CO2H, (CH,)~NH2, CH2C6E~,OH, or CH2C6Hs and Rl is E, CX3, C,~Is, CH2CH=CH2, or CH2C6Hs is constructed p~r the following scheme.

w095128640 2 1 87 792 r l,- 5 2~

O XI~Fmoc HO ~
~OX Cl DEAD. PPh3, PhMe ~EP A
O ~Fnoc ~\o~~\o Cl o II
TEA STEP B
DC~
O Xl~Fmoc (j~H2 ~oJ~
DlC. H03t HO~ Cl DIIF o III
STEP C

W095128640 2 1 8 7 7 ~ 2 r_l" 9! ''^~

IV o HNFmoc ~3 = POLYSTYEE RESIN
1 ) TAGS ~ c ST~P D 2) 20~ PIPERIDIÆ IN Dl~F
FmocN CO
EI F
o ~IN~HNFmOc 1 ) TAGS lXd t 2) 20% PIPERIDINE/DIIF
3~ 5 AcOI~/D~lF

STEP E

wo 95/28640 2 1 8 7 7 9 2 , ~., . ~ ~
-84- Cl Q
o VI
1~ LITHIATED 5(PEhYLNET~YL)-2-OXAZOLIDI~ONE
STEP F T~F, -78 C
2) R X. DMF
X=~3ROIIINE 0~ IODINE
~) TFA:H20:DL~IETHYLSULFLDE
95:5:10 ~'"H ~ ~\
O VII
STEP G hv (350 nm) DIIF

~ W0 95/28640 2 1 8 7 7 9 2 . ~
--8~--.
-! STEP G
Cl~ N0 H0 \~ ~ ~ + -~ H~CX0 ~0 0 vrrr STEP 1~ Ce(NH4)Z(N03)S

~H~ CH0 o : 2 W09s/286~0 1 87792 ~.,. c~

The bPn7Q~ rinP~ VIII are constructed on polystyrene beads similarly to the method of Bunin and Ellman (JACS, ~L. 10997-10998 [1992] ) except that a photolabile linker 5 i8 incorporated between the bead and the bPn7~ 7epine (see steps A, B, and C), thus allowing the benzo~ 7Pr;nP
to be removed in step G non-hydrolytically by exposure to U.V. light ~350 nm in DMF for 10 minutes to 12 hr).
Additionally, binary codes are introduced in steps D and 10 R which allow for a precise flPtPrmln~tion of the reaction seSIuence used to introduce each of the 6 R' s and 5 R~' s .
After removal of the tags according to step H and analysis by electron capture detection ~ollowing GC separation, the nature of the individual R and Rl groups is determined.
1~
Steps D, E, and F PF~5Pnt;~lly follow the ~Lu~eduLe of Bunin and Ellman, but also include the incorporation o~
~Pnt;~iprs IXa-c in step D and IXd-f in Step E. The identi~iers are all represented by Formula IX, Cl ~ le C

IX

~ wo g~o 2 1 8 7 7 ~ 2 wherein:
IX~, indicates n=6;
IXb indicates n=5;
IXC indicates n=4;
5 IXd indicates n=3;
IX~ indicates n=2; and IX~ indicates n=1.

The codes for each of R and Rl are as follows:
Table 4-1 IX R

a CH3 b CH (CH3), a, b CEI2CO2H

c (CH2~ ~NH~

a, c CH~ - C6H~ - 4 - OH

b, c CH,C6Hs IX

d H

2 0 e CH3 d, e C,H5 f CH,CH-CH2 d, f CH2C6Hs Wo sSQ8640 2 1 8 7 7 9 2 ~ ~11.1~ . ~ ~

Ste~ A
To a solution of I (1 equiv) in tolue~e (conc. = 0.5 M) is added the Fmoc protected 2-amino-5-chloro-4 ' -hydroxy-benzophenone (1.3 eq)and diethyla~aodicarboxylate (1.3 eq) 5 and triphenylrhnsrh;np (1.3 eq). The mixture is stirred at room temperature for 24 hr. The solvent i8 removed ~a vacuo and the residue triturated with ether and f iltered and the solvent again removed in vacuo. The resultant product II is purif ied by C~ to~raphy on silica gel .

Step B
To a snll~t;orl of II in DCM (0.2 M) etirring at r.t. is added TFA (3 equiv. ) and the Bolution i6 allowed to stir for 12 hr. The cnl--tinn iB evaporated to dryness in vacuo 15 and the residue dissolved in DCM, washed once with brine and dried (Na2SOf ) . Filtration and evapora~ ion of the solvent affords III. ~ :

steP C
20 19~ DVB (divinylh-~n7~n~) cross-linked polystyrene beads (50~) fllnrtinn~l; 7ed with ~m; - thyl groups (1.1 mE~q/g) are ~llCp~-n~ in DMF in a peptide reaction vessel ~Merrifield vessel) . III (2 equiv) and HOBt (3 equiv) in DMF are added and the vessel shaken for 10 min. DIC (3 eq) 25 is added and the vessel is shaken until a negative Ninhydrin test indicates completion of the reaction after 12 hr.

~ W09~640 21 87792 , , ~

The DMF is removed and the resin washed with additional DMF (x5) and DCM (x5) before drying ;n vacuo.

Stel~ D
5 The dr,y resin is divided into 6 reaction vessels and is 8~ ~p,ont1~d in DCM. The appropriate comb;n~tinn~ of ;d~n~; Ci~rs IX, C (see Tabls 4-1) are added to the flasks and shaken for 1 hr. 'Ille Rh ~TFA) 2 catalyst (1 mol9~) is added to each flask and :haken for an additional 2 hr.
10 The flasks are drained and the resin washed with DCM (x5) .
The resin i8 then treated with a solution of TFA in DCM
(0.01 M) and shaken for 30 min. and then washed again with DCM (x3) followed by DMF (x2). The resin is treated with a 20~ snlllt;nn of piperidine in DMF and shaken for 30 min.
15 and is then washed with DMF (x3 ) and DCM (x3 ) .

To each flask is added the appropriate Fmoc protected amino acylfluoride (3 ~quiv) (when re~uired side-chain fllnrtinn:~l group8 are prote~cted as tert-butyl ester (Asp), 20 te~t-butyl ether (Tyr) or tert-butyloxycarbonyl (Lys) ) with 2,6-di-tert-butyl-4-methylpyridine (10 equiv) and the flasks shaken overnight or until a negative Ninhydrin test is achieved. The resin is washed once (DCM) and then the six batches are ~~ ' in~d and washed again (DCM, x5) before 25 drying in vacuo.

Wo ssnsG40 2 1 8 7 7 9 2 --so The dry resin is divided into five reaction vessels and is suspended in DCM. The appropriate combinations of n~;f;~rg IXd~ (gee Table 4-1) are added to the flasks 5 and shaken for 1 hr. The Rh (TFA) 2 catalyst (1 mol~ is added to each flask and shaken ~or an i:lM~ n~ 2 hr.
The flasks are drained and the resin washed with DCM (x5) .
The resin in then treated with a solution of TFA in DCM
(o . 01 M) and shaken for 30 min. and is then washed with 10 DMF (x3) and DCM (x3) To each flask is added a Rolution of 59~ acetic acid in DMF
and the mixtures are heated to 60C and shaken overnight.
The solvent is drained and then the resin washed with DMF
15 (x5).

Each batch of resin is sl~rPnrl~cl in THF and the flasks are cooled to -78C. To each flask is added a solution of 20 lithiated 5- (phenylmethyl) -2-~ ;nr~n~ (2 equiv) in THF and the mixtures are shaken at -78C for 1 hr. The appropriate alkylating agent (Table ~-2) (~ equiv) is then added to each reaction fla8k followed by a catalytic amount of DMF. The vessels are allowed to warm to ambient 25 temperature and shaken at this t~ ~ t~-re for 5 hrs. The solvent is removed by filtration and the resin washed with THF (xl) and then dried in vacuo~ The batches of resin are then, ` ~n~-l and washed with THF (x2) and DCM (x2) w09sn8640 2 1 8 7 7 9 2 , ", . ~

and the combined resin is then treated with a 95: 5 :10 mixture of TFA:water:dimethylsulphide for 2 hrs to remove the side chain protecting groups.
TAB~E 4-2 lV~;N'l'l~ lt ALKYIIATING
AGENT

e H3CI

d, e C2HsBr f BrCH2-CH=cE2 d, f BrCH2C6Hs ~Ç~
The resultant b~n7o~l;A~ep;n~ can be cleaved from a bead of polystyrene by s~ pen~l;n~ the bead in DMF and irrA~;At;~
with U.V. (350 nm) for 12 hr6.

A bead of interest is placed into a glass capillary tube.
Into the tube is syringed 1 ~L of lM aqueous cerium ( IV) ammonium nitrate (CAN) solution, 1 1l~ of acetonitrile and 20 2~ of hexane. The tube i9 flame sealed and then centrifuged to ensure that the bead is immersed in the reagents. The tube is placed in an ultrasonic bath and ted from 1 to 10 hrs preferably f~om 2 to 6 ~.

w0 9~286410 2 1 8 7 7 9 2 r~

The tube i8 cracked open and ~1 ~L of the upper hexane layer, is mixed with ~0.2 ~L of bis (trimethylsilyl) -t:,m;r~ (BSA) prior to injection into the GC and each tag mem,ber determined using electron capture detection, as 5 ~ _l;fie~ in the following scheme.

Woss/2s640 21 87792 p "1 , O Cl CAN. CH3CN
OMe Cl~-l 2 Cl ,--~ Ho ~o~Cl + lleOll + ClJ~\CI
Cl ~SA
Me Cl lle-Si--O \/Cl~CI
Cl .

wo 95n864o 2 1 8 7 7 q 2 ~ Q ~

I XAMPLE: 5 117. 649 Pe~tide ~ibrarv An encoded library of 117,649 peptides was prepared. This library had the sequence H~N-xxxxxx~ r~r~ -bead~ where the 5 variable residue X was D,E,I,K,I"Q or S. This library was encoded usirg the 18 tags as defined in Example 3; three binary bits being sufficient to represent the seven amino acids used in each step The code was: 001=S; 010=I;
011=R; 100=~; 101=Q; llO=E; and lll=D, where 1 indicates 10 the presence and 0 indicates the absence of a tag.

The constant segment of the library (H7-~RT~nT~r~GG~-bead) was 8ynt~ s; 7~'d on 1.5 g of 50-80 11 Merrifield polystyrene synthesis beads fl~nrt;nn~l;7efl with 1.1 mEq/g of 15 aminomethyl groups using standard solid phase methods based on t-Bu s~ rl~;n protection and Fmoc ~~-;nrh~;n protection. After deprotecting the N-t~rm;n~l Fmoc protecting group with diethylamine, the beads were divided into seven 200 mg portions, each portion being placed into 20 a different Merrifield synthesis vessel mounted on a single wrist-action shaker.

.

The beads in the seven vessels were processed as in Table 3-1 to attach the sets of ;t1Pnt;f;Pr~ (T1-T3) and the 25 corro~pnn~l;nS amino acid to each portion except that instead of DIC, i-butylchlorofo~mate was used for activation .

21 877~2 ~A

This procedure first chemically attached small amounts of appropriate identif iers via their carboxylic acids to the synthesis beads. This attArl was achieved by activating the linker carboxyl groups as mixed carbonic 5 anhydrides using ~Qbutylchloroformate, and then adding an amount of activated; rl~nt; ~; er corresponding to 196 of the free amino groups attached to the beads. Thus, about lg~i of the free amino groups were t.orlr;n~ter~ for each nt;~;er added. The r~ ;n;nr~ free amino groups were lO then couplsd in the usual way with the corresponding protected amino acids activated as their symmetrical anhydrides .

After washing, the seven portions were ' ;nF~d and the 15 Fmoc protected amino groups were deprotected by treatment with diethylamine. The beads were again divided into seven portions and pLu~:~ssed as before, except that the ~-~r~,~,iate identifiers carrying tags T4, T5, and T6 were added to the reaction vessels.
20 The procedure of dividing, labelling, rollrl; nr~ the amino acid ;n;ng and main-chain deprotection was carried out a total of six times using identifiers bearing tags Tl-T18, affording an encoded peptide library of 117,649 25 different members.

w0 9~640 2 1 8 7 7 q 2 --g6-T~ir~l Identifier Preparation To a solution of 8-bromo-1-octanol (0.91 g, 4.35 mmol) and 2,4,6-trichlorophenol (1.03 g, 5.22 mmol) in DMF (5 mL) was added cesium ~ rhnn~te (1.70 g, 5.22 mmol) resulting in the evolution of gas and the precipitation of a white solid. The reaction waa stirred at 80O C for 2 hrs. The mixture was diluted with toluene (50 mL) and poured into a separatory funnel, waslle~=with 0.5 ~ NaOH (2x50 mL), lN
I~Cl (2x50 mL) and water (5~ mL) and the organic phase was dried (MgSO,). Removal of the solvent by evaporation gave 1. 24 g ( 8796 yield) of ta~ a8 a clear oil .
The above tag (0.81 g, 2.5 mmol) was added to a 2 M
solution of phosgene in toluene (15 mL) and stirred at room temperature f or 1 hr . The excess phosgene and the toluene were removed by ev~ror~t i nn and the resulting crude chloroformate was dissolved in DCM (5 mL) and pyridine (0.61 mL, 7.5 Irmol). tert-Butyl 4-hydroxy-methyl-3-niLL~b~.z.,a~e (Barany and AlhPr;rin, J. Am. Chem.
Soc., (1985), 107, 4936-49~2) (0.5 g, 1.98 mmol) was added and the reaction mixture stirred at room temperature f or 3 hrs. The 801~lt;n~ was diluted with ethyl acetate (75 mL) and poured into a separatory funnel. After washing ~ith lN HCl (3x35 mL), saturated NaHCO3 (2x35 mL) and brine (35 mL), the organic phase was dried (MgSO~). The solvent was removed by evaporation and the residue purif ied by chL~ to~raphy on silica gel (5~,; to 7.5% ethyl acetate in ` 21877 095/z8640 92 ~ . s -petroleum ether) affording 0.95 g (79g~ yield) of the ;rlPnt;fier tert-butyl ester as a clear oil.
.
Trifluoroacetic acid (3 mL) was added to a solution of the i~l~nt;f;er tert-butyl ester (0.95 g, 1.57 mmol) in DCM (30 mL) to deprotect the linker acid (i.e., Fl-F2 of Formula I) and the solution was stirred at room temperature for 7 hrs. The mixture was then evaporated to dryness and the residue redissolved in DCM (30 mL). The solution was washed with brine ( '0 mL) and the organic phase dried (MgSO;) . Removal of the solvent by evaporation gave 0 . 75 g (87~ yield) of the i~ont;f;er (6B) as a pale yellow solid. (Tag r rl Ature i8 the game as in Example 3 ) .
Tvpical Encoded Librarv Svnthe8is Step N~-Fmoc-E (tBu) -E (tBu) -D (tBu) -~-G4-Ni~-resin was s~p~n~ cl in DMF (20 mL) and shaken for 2 min. After filtering, 1:1 diethylamine:DMF (40 mL) was added to remove the Fmoc protecting groups and the resin was shaken for l hr. The resin was separated by filtration and washed with DMF
(2x20 mL, 2 min each); 2:l dioxane: water (2x20 mL, 5 min each), DMF (3x20 mL, 2 min each), DCM (3 x 20 mL, 2 min each) then dried in vacuo at 25 C. (The resin was found to have 0 . 4 mmol/g amino groups by picric acid titration at this stage. ) 150 mg Portions of the resin were placed into seven Merrifield vessels and suspended in DCM (5 mL). The W0 9sl28640 2 1 8 7 7 9 2 r~

appropriate identif iers were activated as their acyl carbonates as follows (for the first coupling): T1 (6.6 mg, 0 . 0098 mmol) was dissolved in anhydrous ether (2 mL) and pyridine (10 ~L) was added. Isobutyl chloroformate (1.3 ~L, 0.0096 mmol) was adde~ as a solution i~ anhydrous ether (0.1 mL). The resulting mixture was stirred at 25 C for 1 hr. during which time a fine white precipitate formed. The stirring was sto3?ped and the precipitate was allowed to settle for 30 min. Solutions of the acylcarbonates of T2 and T3 were prepared in the same way.
Aliquots (0.25 mL) of the sllr~rn~t~nt solution of activated identifiers were mixed to give the appropriate 3-bit binary tag codes and the appropriate coding mixtures of ; ~nt.; f; er8 were added to each of the seven synthesis ~Tessels. The vessels were shaken in the dark for 12 hrs, and then each was washed with DCM (4xlO mL, 2 min each).
A solution of the symmetrical anhydride of an N~Y-Fmoc amino acid in DCM (3 equivalents in 10 mL) was then added to the corresponding coded batch of resin and shaken for 20 min. 596 N,N-diisopropylethylamine in DCM (1 mL) was added and the mixture shaken until the resin gave a negative Kaiser test.
~.he resin batches were filtered and combined, and then ~vashed with DCM (4x20 mL, 2 min each), i~ ".~a.~ol (2x20 mL, 2 min each), DCM (4x20 mL, 2 min each) . The next cycle of 1 ~h~l 1; n~/roupl; n~ was initiated by Fmoc deprotection as described above.
-W095/28640 2187792 .~

After Fmoc deprotection of the residues in the last position of the peptide, the side chain functionality was deprotected by sll~r,onA;n~ the resin in DCM (10 mL), adding thioanisole (2 mL), ethanedithiol ( 0 . 5 mL) and tri-5 fluoroacetic acid (10 mL) then shaking for 1 hr at 25 C.
The resin was then washed with DCM (6x20 mL, 2 min each) and dried.

Electron Cal~llre Gas ~'hromatoqra~hv Readinc: of Code 10 A single, selected bead was placed in a Pyrex capillary tube and washed with DMF (5xlO ~LIJ). The bead was then suspended in DMF (1 ~LL) and the capillary was sealed. The suspended bead was irradiated at 366 nm for 3 hrs to release the tag ~1 ~nhnl ~, and the capillary tube 15 subsequently placed in a sand bath at 90 C for 2 hrs.
The tube was opened and big-trimethylsilyl iqnet:~m;~ (0.1 mL) was added to trimethylsilylate the tag alcohols.
After centrifuging for 2 min., the tag solution above the bead (1 ~L~) was injected directly into an electron capture 20 detection, ~ri 1 l P~ry gas cll, to~raph for analysis . Gas c~ ~, tn~raphy was performed using a Hewlett Packard Series II Model 5890 gas chromatograph equipped with a 0.2 mmx20 m methyl ~;1; t-nn~ fused silica capillary column and an electro~ capture detector. Photolysis reactions were 25 performed using a UVP "Black Ray" model WL 56 hand-held 366 nm lamp.

wo ss/2s640 ~ 2 1 8 7 7 q 2 r ~ I I

Ant; hn~y Af f initv Methods The anti-C-myc peptide monoclonal antibody 9E10 wa6 prepared from ascites fluid as described in Evans et 3~, Mol. Cell Biol, 5, 3610-3616 tl985) and Munro and Pelham, Cell, 48, 899-907 ~1987). To test beads for binding to 93310, beads were inrllhAt~rl in TBST [20 mM Tris-HCl (pH

7.5), 500 mM NaCl and 0.05% Tween-20] rnntA;n;n3 1% bovine serum albumin ~BSA) to block non-specific protein binding sites. The beads were then centrifuged, resuspended in a 1:200 ~ lt;nn of 9E10 ascites fluid in TBST + 1% BSA and incubated overnight at 4C. Beads were subsequently washed three times in TBST and ; ncllhAt~d for 90 min. at room t: ~ -rAtllre in AlkAl ;n~ phosphatase-coupled goat ~nt; ~e IgG Ant;hofl;r~ (Bio-Rad Laboratories), diluted 1:3000 in TBST + 1% BSA. After washing the beads twice in TBST and once in rhnsrhAt~Re buffer (100 mM Tris-HCl, pH
9.5, 100 mM NaCl and 5 mM MgCl2), the beads were incubated 1 hr at room temperature in phosphatase buffer rnntA;n;n~
one one-hundreth part each of AP Color Reagents A & B
(Bio-Rad l,aboratories). To stop the reaction, the beads were washed twice in 20 mM sodium EDTA, pH 7.4. Solution phase affinities between 9E10 and various peptides were det~rm;nPcl by a modification of the competitive El,TSA
assay described by Harlow et al., ~nt; horl;~ a Iaboratory Manual, 570-573, Cold Spring Harbor Press, Cold Spring Harbor, N . Y .

wo9s~640 2 1 8 7 792 P~

From a 30 mg sample of the combinatorial library of peptides, 40 individual beads were i~lPn~;f;ed which stained on exposure to the anti-C-myc monoclonal antibody.
3PrQ~;n~ of thege positive-reacting beads established the 5 ligand~s reaction se~uence as the myc epitope (T~QT~TlT.~nI~) or sP~1PnrPC that differed by one or two substituents among the three N-terminal residues.

23 . 540 . 625 Mixed ~n~; de ~ibrarv The PnrQ~; n~ techni~aue was tested further by the preparation of a, ' ;n~torial library of 23,540,625 members consisting of peptides and other amide ~ ~ullds.

The synthesis was carried out using 15 different reagents in 5 steps and 31 different reagents in the sixth step.
Four ;r~Qnt;f;Prs were used to encode each of the 5 steps with 15 reagents and five i~pnt;f;ers were used in the final step with 31 reagents. A label set of 25 i~Pn~;fiprs was therefore ~Le~a-cd. 2-Nitro-4-ca-L~,~yl,~ yl, O-aryl substituted ~I1-hydroxyalkyl r~rhr~n~tP
identif iers were employed, where the tag , ~ c were comprised of an alkyl moiety of from 3 to 12 carbon atoms and the aryl moieties were (A) pentachlorophenyl, (B) 2,4,5-trichlorophenyl, (C) 2,4,6-trichlorophenyl, or (D) 2,6-dichloro-4-fluorophenyl. A 3et of 25 tags was prepared using appropriate alkyl chains lengths with A, B, C or D, separable using a 0 . 2 mMx25M methylsilicone GC
.. _ _ _ _ _ _ _ . _ _ _ _ _ .. .. _ .. , ~ . . _ _ W09sl28640 '~ C -column. The chemical compositions of tags T1=T25 (where T1 represents the tag with the longest retention time, and T25 the tag with the shortest retention time) are summarized below-5 T1 lOA T6 lOC T11 7B T16 5C T21 2B

T38A T8 9C T13 6B Tl8 4C T23 lB
T47A T9 8B T14 6C Tl9 3B T24 lC
T5lOB T10 8C T15 5B T20 3C T25 2D

The desi~n~ nc lOA, 9A, etc. are as described in Bxample 3.

The ~i~teen reagents used in the first five stages and the 15 code identi~ying them are represented below where 1 ~ep~e~ent~ the pre~ence of tag and ~ the ab~en~e thereof.

wo 95/28640 2 1 8 7 7 9 2 1 ~ ., , ~

REAGENT CODE
- L-serine (0001) D-~erine (0010) ~-glutamic acid (0011) 5 D-glutam~ c acid (0100) L-gluLamine (0101) D-glutzmine (0110) ~- ly3ir~ ( 0111 ) D-lysine (1000) 10 L-Proline (1001) D-Proline ( 1010 ) L-phenyl~ nln~ (1011) D -phenyl a;~.nine ( 110 0 ) 3 -amino -~enzoic ( 1101 ) 15 acid 4-~-n;n~ rh~nyl (1110) acetic acid 3, 5-diamino- (1111) benzoic acid ~ r W0 95/28640 ` ` ' 2 1 8 7 7 9 2 P ~
~ 104 -The 31 reagents and the code representing them in the 6ixth stage are represented below:
REAGENT CODE
l.-serine (00001) 5D-serine (00010) JJ-glutamic acid (00011) D-glutamic acid (00100) L-gl ~ m; nf~ ( 00101) D - glut amine ( 0 0110 ) ~-lysine (00111) D-lysine (01000) L-proline ~01001) D-proline (01010) L-phenyl~l ~n; nP (01011) D-phenyl;-l~n;n~ (01100) 3-amino-benzoic acid (01101) 4 _;~m; nmrhPnyl acetic acid ( 01110 ) 3,5-diamino-benzoic acid (01111) Succinic Anhydride (10000) Tiglic acid (10001) 2-pyrazine carboxylic acid (10010) Wo 9v/~8640 2 1 8 7 7 9 2 rv~

(~) thioctic acid (10011) 1-piperidinepropionic acid (10100) piperonylic acid (10101~

6-methylnicotinic acid (10110) 3- (2-thienyl) acrylic iacid (10111) methyl iodide (11000) tosyl chloride (11001) p-tol-l~n~clll fonyl isocyanate (ilO10) 3-,vy~ 70ic acid (11011) 10 phth;ll 1; c anhydride (11100) acetic anhydride (11101) ethyl chlorof ormate ( 11110 ) mesylchloride (11111) 15 A spacer of six glycine units wa3 ~le:~ale:d on the beads using standard methods. The variable region was constructed using butyl 8i'lf'rh~; n protection, and amino groups were protected as Fmoc derivatives. Amide bonds were formed by activation of the carboxylic acid with DIC

2 O and ~IOBt .

Wo 9S/286~l0 ',; ~ , ~ i ``, 2 1 8 7 7 9 2 EXANPLE: 7 Hetero-Diels-~lder I,ibrarv A combinatorial hetero Diels-Alder library comprising 42 I '- of the formula:

Ar ~S~R
o H
wherein;
Rl is H, CH30, F3C, F3C0, HsC60~ or C6H~l;
R' i s H, C~I3, or CH30;
15 R3 is H (when n=2), or CH3 ~when n=1); and ~0~ R = ~ or Cl or Ar =

~ wo gs/28640 2 1 8 7 7 9 2 P~ll~ ' ''^~

was constructed per the followiIlg scheme:
O2~ HO~
~OH
DEAD, PPh3, Ph~l~
I STEP A
2~
~0~~ R

~FA STEP B
DCil 1) (~H2 2 DlC, 80~t, DhF ~ O~
2) kfentifiers Xab HO o R
ST~P C III

.

WO 95/28640 t. ~ 2 1 8 7 7 9 2 STEe C Xa.b~~

IV
1) Toluene 2) Identifiersxc-e STEP D
.b R2 --` H~ o~N~ R
~c e OzN V R R
1) Identif~ers X~
2) [~\o R
BF3. Et20 DCn STEP E

wo gs~2864~) 2 1 8 7 7 9 2 , ~

4 b ~ ~, ~R
21~ ,, R ~1 hv ~350 mn) Dh'F
STEP F
~,~ g ~CHO HO~R

5~ ; Ce(NH4)2(NO3)6 N~` CHO
NO

Wo 95128r~0 ~ ,r~ s i 2 1 8 7 7 9 2 ~ J~ c Ir ~ ~

The azatricyclic products tVI) were constructed on polyYLyLe:lle beads and were linked to the beads by a photocleavable linker allowing the azatricycle (VII) to be removed from the bead by exposure to ~.V. light (350 nm in 5 DMF) . The binary codes introduced in steps C, D and E
allow a unique determination of the reaction sequence used to introduce Ar~, Rl, R2 ~.nd R3. The encoding tags were removed according to 5 cep G and analyzed by electron capture detection follo~ g GC separation.

The identifiers used in ~his scheme are represented by the formula X:
Cl ~:
o Wherein;
20 X~ indicates n=10 Xb indicates n=9 Xc indicates n=8 Xd indicates n=7 X~ indicates n=6 25 X~ indicates n=5 X2; nl~ t.o~l n=4 The codes for each of R, Rl, R', R3 are as follows:

2 1 8 7 7 9 2 /L~.
Wo 95n8640 1 ~ ' TAsLE 7-1 s 5 a Ar = ~ R = H
b Ar = J~ R = Cl a,b Ar = ~OH
C Rl=H R2=H
d Rl=H R2=CH3 d, c Rl=OCH3 R2=OCH3 e Rl=CF3 R2~H
e, c Rl=C6hsO R2=H
e, d Rl~sF3Co R2sH

wO95n8640 ' 2187792 r ~

e,d,c R~=C6Hll R'=H

f R3=CH3 n=1 g R3=H n=2 SteP A
To a solution of I (2 . 03 g, 8 mmol), 4 -hyu~ u~yb~ Aphyde ~1.17 g, 9.6 mmol) and triphenylrh~1sphinp (2.73 g, 10.4 mmol) in toluene (20 mL) stirring at 0C was added over a 10 period of 3 0 minutes diethylazodicarboxylate . The solution was allowed to warm and stirred for 1 hour once ambient temperature had been reached. The solution was cr~nf~ntrAt~d by removal of approximately half of the solvent ~ vacuo and was then triturated with ether. The 15 mixture was then f iltered and the residue was washed thoroughly with ether. The solvent was removed n vacuo and the residue was purified by chromatography on silica g~el (159~ ethyl acetate in hexane) affording 1.3 g of the ether IIa (479~ yield).

2-chloro-4-hydroxybenzaldehyde and 2-hydroxy-1-nArhth:~1 dehyde were coupled to I in An~ go1l~ fashion a~ording ethers IIb and c in yields of 91% and ~7~6, respectively .

SteP B
To a solllt;~)n of ether IIa (0.407 g, 1.14 mmol) in DCM (20 mL) stirring at room temperature was added TFA (8 mL).

wossns640 2187792 P~l/. 5 - ~

The solution was allowed to stir for 6 hrs. The solution was evaporated to dryness i vacuo affording 0.343 g of acid IIIa (lOOg6 yield). Ethers IIb and IIc were deprotected analogously affording acids IIIb and c in yields of 92~ and 1009~ respectively.
,. , SteP C
Into a peptide reaction ves6el (Merrifield vessel) were measured lY6 DVB (divinylbenzene) cross-linked polystyrene beads (50-80~) functinn~l;ze-l with ~m; - 'hyl groups (1.1 meq/g) (200 mg of resin). The resin was suspended in DMF
(2 mL) and shaken for 20 min. The acid IIIa (38 mg, 2 equiv. ), 1-1lydl ~,~yl,ellzotriazole (40 mg, 2 equiv) and diisopropyln~rhn~l;;m;~lP (38 mg, 2 equiv) were added and 15 the mixture ahaken until a negative Ninhydrin test was achieved (22 hr). The c~ tinn was removed by filtration and the resin was washed with DCM (8x 10 mL).
The resin was rpcl~qr~n~pd in DCM (5 mL), identifier Xa (15 20 mg) was added and the flask was shaken for 1 hr- Rh (TFA) 2 catalyst (1 mol~) was added and the flasks shaken for 2 hrs. The solvent was removed by filtration and the resin rPR~lcp~nrlP~ in DCM (5 mL). Trifluoroacetic acid (1 drop) was added and the vessel shaken for 20 min. The solvent 25 was removed by filtration, and the resin was washed with DCM (8x 10 mL).

W0 95128640 ~ ~ 2 1 8 7 7 9 2 In an analogous fashion, acids IIIb and IIIc were attached to the resin and were encoded with the appropriate i~l.ontif;er3, i.e., Xb for acid IIIb and Xa and Xb for acid IIIc. The three batches of resin were combined, mixed, 5 washed, and dried.

The dry resin was divided into 7 equal portions ( 87 mg) which were put into seven peptide reaction vessels 10 (Merrifield vessels) which were wrapped with heat tape.
The resin in each vessel was suspended in toluene (10 mL) and shaken for 20 min. An appropriate amount of one aniline was then added to each flask (see Table 7-2).

TA~3LE 7-2 FI.A8~ ANILINE AlIO~r ADDED
Aniline 3 mL
2 3, 5-dimethyl ::ln; 1; nf~ 3 mL
3 3, 4, 5-trimethoxyaniline 2 g 20 4 4-trifluo~, ~hy1;3n;1;n~ 3 mL
5 4-phenoxyaniline 2 g 6 4-triflu.,~ h~Yyaniline 3 mL
7 4-cyclohexyl~n;l;n., 2 g The heating tape was connected and the reaction mixtures shaken at 70C for 18 hrs. The heat tape was ~;q,rmn-~ct~ and the solvent was removed by filtration and each batch of resin was washed with dry DCM (4x 10 mL), ~ther (10 mL), toluene (10 mL) and DCM (2x 10 mL). Each 30 of the portions was then s-l~p~n~lPd in DCM (5 mL) and to ~each f lask was added the appropriate identif ier or W095/28640 21 ~7792 r~

in~t~n~ of identifiers (Xc-e) (15 mg) (see Table 7-1) .
The flasks were shaken for 1 hr. and then Rh(TFA)2 (1 mol~) was added to each flask and shaking ~nnt;nllPd for 2 hrs.
5 The solvent was then removed and each batch of resin was re-suspended in DCM (5 mL) to which was added TFA (1 drop). This mixture was shaken for 20 min., then the solvent was removed by f iltration . The batches of resin were then washed (DCM, lx 10 mL) and c~ ' ;nP~I, washed 10 again with DCM (3x 10 mL) and then dried thoroughly ' n yacuo .
Ste~ E
The dried resin was divided into two e~ual portions ( 0 . 3 15 g) and each was placed in a peptide reaction vessel. The resin batches were washed with DCM (2x 10 mI,) and then resuspended in DCM (5 mL). To one flask was added the irlPnt;f;Pr Xf (15 mg) and to the other was added Xg ~15 mg). The flasks were shaken for 1 hr. prior to the 20 addition of Rh(TFA)2 catalyst (1 mol9~). The flasks were shaken for 2 hrs. and then the solvent was removed hy filtration. Each batch of resin was washed with DCM (3x 10 mL), and each was then rpcll~ppn~lpd in DCM (5 mL) .
25 The appropriate enol ether (1 mL) (see Table 7-1) was added to the flasks and the vessels shaken for 30 min. To each flagk was added a solution of BF3-0Et2 (0 . 5 mL of a 59 solution in DCM) and the flasks were ~haken for 24 hrs.
Removal of the solvent by filtration was followed by 30 washing of the resin with DCM (10 mL) and the resin was then ~: ' ;nPd. The beads were then washed further with DCM (5x 10 mL), DMF (2x 10 mL) h:lnnl (2x 10 mL) and DCM
(2x 10 mL) . The resin was then dried thoroughly ~.~L vacuo.
~.
35 Stel~ F
To conf irm the identity of the products produced in the Hetero-Diels-Alder library one example was completed on a .. . . ~

~09sl28640 ; ' 21 87792 p large scale to allow confirmation of the structure by spectroscopic means. The procedure followed was P~5Pn~ l ly the game method as described for = the c in~nl-ial library. In step A 4-llydl~yl.PIl7~ Phyde 5 was coupled to the photolabile group. In step D, aniline was t ~n~Pn~P~ with the aldehyde. In step E, the enol ether was formed with 4,5-dihydro-2-methylfuran.
The photolysis of the compound (step F) was performed by ,~llqpPn~l;n~ 100 my of the beadg in DMF (0.3 mL) and irr~-l; At; n~ the beads with WP "Black Ray" model UVL 56 hand-held 366 nm lamp for 16 hrs. The DMF was removed to one side by pipette and the beads rinsed with additional DMF (2x 3 mL) . The ~ig;n~l golution and the wlflh;n~c 15 were combined and the solvent removed ~ aQ- NMR
analysis of the reaction mixture showed it to contain the desired azatricycle by comparison to the authentic sample.
Step G
20 A bead of interest was placed into a pyrex glass capillary tube sealed at one end. A solution (1 ILL) of lM aqueous cerium (IV) ammonium nitrate and acetonitrile (1:1) was syringed into the tube, and the tube was then centrif uged 80 that the bead lay on the bottom of the capillary and 25 was completely ; ~ ~ by the reagent solution . Hexane (2 ~L) was added by syringe and the tube was again centrifuged. The open end of the capillary was flame-sealed and placed in an ultrasonic bath for 4 hrs. The ry wa8 then placed inverted into a centrifuge and 3 0 spun such that the aqueous layer was f orced through the hexane layer to the bottom of the tube. This extraction process was repeated 3 or 4 times and the tube was then opened. The hexane layer (1.5 ~L) was removed by syringe and placed int,o a different capillary crm~;n;n~ BSA (0.2 35 ~L). This tube was sealed and centrifuged until the reagents were thoroughly mixed. A portion of the solution (ca. 1 ~L) was removed and injected into a gas .

w095/28640 2187792 ~ [-~

to~raphy machine with a 25M x 0.2 mM methylsilicone fused silica column with Pl~ tr~n capture detection for separation and interpretation of the tag molecules.
J 5 The sample was injected onto the GC column at 200C and 25 p5i of carrier gas (He2). After 1 minute the temperature was increased at a rate of 20C per minute to 320C, and the pressure was increased at a rate of 2 psi per minute to 40 psi. These conditions are shown in the following diagram:
GC COI~DITIOl~S
TEIIPERATURE

/ C per mln 1 min P~ESS~RE ~ 40 psi _/ ' "' .. .
25 psi 1 min wog5n8~0 ' 2 1 87 792 L~

The following results were obtained with four randomly selected beads:
TAG DETECTED
Xf Xe Xd Xc Xb Xa Ar 2-Hydroxy naphthyl Rl C6 R3 CH3 (n=l) TAG J~ ~lL.l) Xg Xe Xd Xc Xb Ar 2-chloro-4-;lydL~,~y~henyl R3 H (n=2 ) Bead 3 TAG l ~
Xg Xe Xd Xb Xa Ar 2-Hydroxy naphthyl R~ H
25R3 H (n=2 ) ,~3ead 4 TAG nr., r.l ' ~
X~ Xe Xd Xb 30Ar 2-chloro-4-l~ydL~,Ly~.he.lyl R3 CH3 (n=1) w09~640 2 1 ~7792 .~.,. "~

E~LJ5 8 Benzo~ 7e~ine LibrarY
-Following the procedure of Example 4, a combinatorial 5 library is constructed of the Formula X

~R4 ~ ~ R

R

wherein 20 R is a radical of a naturally occurring D or L amino acid;
R1 is H, C1-C6 alkyl, lower alkenyl, C1-C6 alkylamine, carboxy C1-C6 alkyl, or phenyl C1-C6 alkyl wherein the phenyl is optionally substituted by lower alkyl, F, Cl, Br, OH, NH2, CO2H, or O-lower alkyl;
25 R2 is H or CO2H;
R3 is H or OH;
R~ is H or Cl;
with the provisos that when R3 is OH, R2 is H and when R2 is carboxy, R3 is H.

.

This library is released from a plurality of encoded beads of the general f ormula n~ 1' FZ~N_ R
~--~0 R

wherein IX,, is a plurality of identifiers of the Formula Ia wherein said plurality Le~ sc:1~ts an encoded scheme;
S is a substrate;
Fl'-F2 is the residue of the linker member of Formula Ia;
and R, Rl, R2, and R' are as defined for Formula X.

~ wo gs/28640 2 1 8 7 7 9 2 EXAMPLE: 9 Tvl~ical Identifier Prel~arations The diazo compound identif iers which are attached to the resin via carbene formation are pL I~CLLed as exemplified.

d9 of the general formula ~C~ ~ ~Ar wherein n is 0-10 and Ar is pPnt~~hlnrophenol, 2,4,6-trichlorophenol, 2,4,5-trichlorophenol, or2,6-dichloro-4-fluorophenol are ~Le~,aLed as follows.
To a solution of l-hydroxy-4-~2,6-dichloro-4-fluoro-phenoxy)butane (O.38 g, 1.5 mmol), isovanillin (O.228 g, 1.5 mmol) and triphenylphnsph;np (0.393 g, 1.5 mmol) in THF (8 mL) was added diethylazodicarboxylate (0.287 g, 1.7 mmol). The solution stirred at r.t. for 36 hrs. The solvent was removed vacuQ and the residue purif ied by C}~L~ tn~aphy on 6ilica gel (with a mixture of 20~ ethyl acetate and 80% petrQleum ether) affording 0.45 g of the aldehyde ( 77% yield) .
The aldehyde (100 mg, 0.26 mmol) was dissolved in acetone (8 mL) and was treated with a ~nlllt;nn of KMnO~ (61 mg, 0.39 mmol) in acetone (4 mL) and water ~4 mL). The reaction stirred at room temperature f or 13 hrs . The mixture was diluted with ethyl acetate (100 mL) and water 2 1 ~ 7 7 9 2 r ~ I L A ~ ~

(50 mL) and the layers were separated. The aqueous layer was extracted with additional ethyl acetate (2x 100 mJJ).
The combined organic layers were washed with water ~50 mL) and dried (MgS04). Removal of the solvent afforded 109 mg of the benzoic acid (93% yield~.
A solution of the acid (76 mg, 0.188 mmol) in methylene chloride (2 mL) was treated with oxalylchloride (36 mg, 0 28 mmol) and catalytic DMF. After stirring for 10 min at room t ~-rAture slow but steady evolution of gas was observed. Stirring cnn~;n~ l for 2 hrs. when the solution was diluted with DCM (15 m~) and washed with saturated aqueous sodium hydrogencarbonate solution (5 mL). The layers were separated. The organic layer was dried (Na2S0~) and the solvent evaporated affording the benzoyl chloride as pale yellow crystals.
The benzoyl chloride was dissolved in methylene chloride (5 ml~) and was added to a stirring Rnl~tinn of an excess of ~11A- 1-hAnP in ether at -78C The cold bath was allowed to warm up and the mixture allowed to stir for 5 hrs at room t~ LuLe. The solvents and excess diazomethane were removed l vacuo and the residue purified by C1-L~ tcgraphy on silica gel using gradient elution method where the concerLtration of ethyl acetate ranged from 1096 to 40 9~ in hexanes affording 48 mg of the diazo .~ ~ ~1 (6096 yield).

~ W0 95n8640 2 1 8 7 7 9 2 1 _l/L_ ~460~

r~ o1ln~ of the general formula:
-~o~le o~Ar O

wherein;
15 n is 0-10 and Ar i~ p~ntArhlnrophenol, 2,4,6-trichlorophenol, 2,4,5-trichlorophenol, or 2, 6-dichloro-4-fluorophenol are prepared as follows.
Methyl vAn;11A~ (0.729 g, 4.0 mmole), 1-hydroxy-9-(2,3,4,5,6-pentachlu,u~henu~y)nonane (1.634 g, 4.0 mmole) and triphenyl~hnsr~ln~ (1.259 g, 4.8 mmole) were dis~olved in 20 mL dry toluene under argon. DEAD (0.76 mL, 0.836 g, 4.8 mmole) was added dropwise, and the mixture was stirred at 25C for one hour. The solution was ~nn~nt~ated to half volume and purified by flash chromatography eluting with DCM to give 1. O g (1. 7 mmole, 439~) of the product as a white crystalline solid.
The methyl ester above (1.0 g, 1.7 mmole) was dissolved in 50 mL THF, 2 mL water was added followed by lithium hydroxide (1.2 g, 50 mmole). The mixture was stirred at 25C for one hour then refluxed for five hours. After cooling to 25(~ the mixture was poured onto ethyl acetate (200 mL) and the solution wa~3 washed with 1 M HCl ~50 mL
x3) then sat. aq. NaCl (lx 50 mL) and dried over sodium W095128~0 ~ i; 2 1 877q2 r~ 5 0l~

~3ulfate. The solvent was removed and the crude acid azeotroped once with toluene.
The crude material above waa dissolved in lO0 ml~ toluene, 10 mL ~1. 63 g, 14 mmole) thionyl chloride was added, and the mixture was refluxed for 90 min. The volume of the solution was reduced to approximately 3 0 m~ by dist;ll~;nn, then the .. ;n;ng toluene removed by evaporation ~ vacuo. The crude acid chloride was dissolved in 20 m~ dry DCM ~nd cooled to -78C under argon and a solution of approximately lO mmole diazomethane in 50 mL anhydrous ether wa~ ~dded. The mixture was warmed to room temperature and s~irred for 90 min. Argon was bubbled through the ~oluti~II for 10 min. then the solvents were removed by evaporation in vacuo and the crude material was purified by flash chromatography eluting with 10-20% ethyl acetate in hexane. The ~ 7nk~tnno (0.85 g, 1.4 mmole, 82% over three steps) was obtained as a pale ~rellow solid.
The following iri~nt;f;-~rs have been prepared as described above:
P~otol ~hi ] e Cleav~ e 50 I~-~n1-;f;~rs were prepared of the formula:
H0 ~O ~ 0--Ar wherein:
Ar is cl l! Cl C~ Cl~[~CI Cl~ F~Cl Cl ~ Cl Cl Cl ~ Wo 95/28C40 2 1 8 7 7 9 2 r and n is 1,2,3,4,5,6,7,8,9, and 10.
Oxidative Cleav~e ^~e I
5 7 T~l~n~;~lers were prepared of the formula wherein:
Ar i8 Cl ;~
and n is 4,5,6,7,8,9, and 10.
Oxidative Cleav^~e T~e II
13 Identifiers were prepared of the formula ~ C:13 0--Ar 3 0 wherein:
Ar is Cl $~
Cl .

. 21 8779 W0 9s/28640 and n is 1,2,3,4,5,6,7,8,9,10;
and wherein:
3Ar is F ~Cl and n i8 0, 3, and 9 . C

~ W09~n8640 2t87792 1~,I/L '~

Example 10 ~nco~; n~ Combinatorial Libraries with Taqs Readable bv Mass Sl~e~l ~.,s~v~ v The tag6 4, 11 and 13 ~Scheme 8) of the same structure, 5 but different molecular weights due to varying deuterium substitution, were each synthesized (Schemes 9 and 10) and separately analyzed by mass spectroscopy (MS). Among MS
techniques, positive chemical ioni~ation mass spectroscopy (PCIMS) gave minimal fragmenta~ion of the tag, such that 10 only the ll.or~ll Ar ion ( [M+N~] ~) and one other fragment ( [MH-H,O] ) were evident (~igures 1, 2 and 3) . This actually allowed the presence or absence of a tag to be determined by the observation of two signals, which removes any possible amhiguity when analyzing a more 15 complex mixture. Apprn~;r-t.oly equal amounts of the t_ree tags were then mixed and analyzed by PCIMS ( f igure 5 ) .
Again, the two signals corrf~qpnntl;n~ to each separate tag could easily be distinguished.
20 Tag 4 was now transformed into the diazoketone precursor

8 (Scheme 9), then attached to Tentagel resin as 9 (Scheme 12). One bead of the 9 was qllhqPqll~ntly removed and 4 oxidatively released using ceric ~ n nitrate. PCIMS
analysis again clearly showed the presence of tag 4.
In summary, the set of tags 4, 11 and 13 of the same structure, but different molecular weights were 8yn~h,~q; ~o~. All were eagily detected simultaneously in a mixture by PCIMS. The small amount of 4 released from 3 0 a single bead of Tentagel resin used in combinatorial library synthesis is detectable by PCIMS. MS is a viable and sensitive detection method for tags, and can be used as the basis for an ~n~ofi;n~ scheme of a combinatorial library.
Analysis of 4, 11 and 13 by PCIMS was obtained using a reagent gas mixture of 1~ ~;H3 in CH,.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ , . .

W0 95l28640 (2) . To a solution of 11.1 mL (125 mmole, 5.00 eq.) of 1,4-but~n~l;ol ~1), 6.97 mL (50.0 mmole, 2.00 eq.) of Et,N
and 0.153 g (1.25 mmole, 0.05 eq.) of 4-dimethylaminopyridine in dry CH2Cl2 (100 mL) at 0C under Ar, was added 3.88g ~25.0 mmole, 1.00 eq.) of 97% tert-butyldimethylsilyl chloride. The resulting solution was stirred at 0C for 15 min, then 25C for 16hours. The reaction was then diluted with CH2Cl2 (250 mL) and washed with 1 M HCl (lOOmL), s~t--r~cl aq. NaHCO3 (100 mL) an~ H20 (100 mL), then dried (MgSO~) . Removal of the volatilee in vacuo gave the crude product 2 as an oil.
t3). To a solution of -10.0 mmole of crude alcohol 2, 1.93 g (10 . 5 mmole, 1. 05 eq. ) of pentafluorophenol and 2 . 89 g (11.0 mmole, 1.10 eq.) of triphenylphosphine ln dry CH2Cl2 (40 mL) at 0C under Ar, was added 1.73 mL (11.0 mmole, 1.10 eq. ) of diethyl azodicarboxylate. The resulting orange solution was stirred at 0 C for 5 min, then 25C
f or 15 hours . The reaction was then diluted with CH2C12 (250 mL) and washed with saturated aq. Na2CO3 (100 mL), saturated ag. NH~Cl (lOOmL) and ~120 (100 mL), then dried (MgSO,) . Removal of the volatiles in vacuo and purification by flash .~._ to~raphy (0-20~ EtOAc in hexanes) gave the product 3 as an oil.
(4). To a sol~lt~n of 1.85 g (5.00 mmole, 1.00 eq.) of silyl-protected alcohol 3 in THF (20 mL) at 25C, was added 10 . O ML (10 . O mmole, 2 . 00 eq. ) of a 1. O M solution of tetrabutylammonium f luoride in THF . The resulting orange solution was stirred at 25C for 4 hours. Removal of the volPt;l~R in vacuo and purification by flash cll., to~raphy (20-409~ EtOAc in hexanes) gave 1.10 g (8696) of the product 4 as an oil_ (5) . To a solution of O . 800 g (3 .125 mmole, 1. 00 eq . ) of alcohol 4, 0.569 g (3.125 mmole, 1.00 eq.) of methyl v~n;ll~t.o and 0.984 g (3.75 mmole, 1.20 eq.) of ~ Wo 95/28C40 2 1 8 7 7 9 2 triphenylrhn6ph;nP in dry CH2Cl2 (20 mL) at 0C under Ar, was added 0.591 mL (3 .75 mmole, 1.20 eq. ) of diethyl azodicarboxylate. The resulting pale yellow solution was stirred at 0C for 5 min, then 25C for 19 hours. The 5 reaction was then diluted with CH2Cl, (100 mL) and washed with 1 M NaOH (50 mL), saturated aq- NH4Cl (50 mL) and ~2 (50 mL), then dried (MgSO~). Removal of the V~lRt;lPR in vacuo and purification by flash ~ tngraphy (209~ EtOAc in hexanes) gave the product 5 as an oil.
(6). To a solution of 3.125 mmole of ester 5 in TH~ (12 mL) was added 1.31 g (31.3 mmole, 10.0 eq.) of lithium hydroxide monohydrate. MeOH (24 mL) was added to the resulting suspension to form a solution, which was stirred 15 at 25C for 1 hours, then refluxed for 1 day. Volatiles were removed in vacuo, and l M HCl then added until solution was -pH l. The white prer;ritRte of product which formed was collected and dried to give 0.968 g (769 - 2 steps) of 6.
(7). To 0.968 g (2.38 mmole, 1.00 eq.) of carboxylic acid 6 under Ar, was added 2.43 mL of thionyl chloride. The resulting s~pon~;or~ was refluxed for 1.5 hours, after which time a yellow snlut;nn had formed. Volatiles were removed in vacuo, and the resulting residue azeotroped three times with toluene to give the product 7 as colorless crystals.
(8). To a ~ol~t;nn of 2.38 mmole of acid chloride 7 in 1:1 THF:MeCN (20 mL) at 0C under Ar, was added 1.16 mL (8.33 mmole, 3.50 eq.) of Et3N followed by 3.57 m~ (7.14 mmole, 3 . 00 eq. ) of a 2 . 0 M solution of (trimethylsilyl) diazomethane in hexanes . The resulting yellow solution was stirred at 0C for lh, then 25C for 35 l day. The reaction was diluted with EtOAc (150 mL) and washed with saturated aq. NaHCO3 (2 x 75 mL) and saturated aq. NaCl (2 x 75 mL), then dried (MgS0j) . Removal of the volatile~3 in vacuo gave the crude product 8 as an oiL.
(11) . Using commercially available 1,4-butanediol-2,2,3,3-5 d, (10) in place of 1, and an analogouc procedure to that described for the trans~ormation 1 into 4, 11 was obtained in 419~ yield over three steps.
Using commercially available ll4-h~lt~np~;rl-2r2~3~3~4r4-d8 10 (10) in place of 1, and an analogous procedure to that described ior the transformation 1 into 4, 13 wa-c obtained in 4296 yield over three steps.
Tag 4 was introduced onto the solid support using 5 to 5096 15 (w/w vs. resin) precursor diazoketone 8 to give 9 by essentially the same procedure given in Example 4; the Hetero Diels-Alder library. Tag 4 was also subsequently removed from 9 by essentially the same procedure as Step G in Example 4.
The ~ 7rkrtr)npc corr~cpnn~ltn~ to tags 11 and 13 are used to introduce these tags onto the solid support, so that along with 8, they yield member~ of a binary ~nro~llnr~ set.

Wo 95128640 2 1 8 7 7 9 2 Scheme 8 Ho~ O~F
r ~ F
If. l = 256 D D F
i/
il 11.~. = 260 D D D D F
HO~0~F
D D D D F/~F

. ~. = 264 . ! I 2 1 8 7 7 9 2 wo 95n8640 -1 3 2~
s~ ,, _ o ~ ,," o ~
~ o --~ _ eu,, _ ~ ~
e ~ 0 ~
f:~ o O o ~i o~

Scheme 10 D D D D F
~ ''~><~~ ~ ~ ~o ~X~ 0 ~ F
D D D D F~F

D D D D D D D D F
---- ~o~O~F
D D D D D D D DF,~ `F

wo95n8640 ` 21 8 7 792 r ~ c~

o~

~3 ~
~,~ D o n~
~. ~

O ~ O~
~1 o ~0 0~

Wo 95/28640 2 1 ~ 7 7 9 2 Pre~aration of Deut~riated Tac~
C02Me C02Me ¢~CD3CD20H
PPh3, D~AD

To a solution o~ PPh3 (3.2 g, 12.1 mmol) in THF (30 mL) at 0 C was added diethyl azodicarboxylate (1.9 ml, 12 mmol).
After 10 minutes at O C, a solution of methyl 4-llyd,~,~y~he~lylacetate (1.66 g, 10 mmol) and d5-ethanol (800 /~L, 12.3 mmol) in 10 mL THF was added. After stirring at 25 C for 2 hours, the reaction mixture was cr~nr~ntrated 15 and purified by flash :11LI to~raphy to give methyl 4-d5 ethoxyphenyl ~ret~te (1. 86 g, 93~) as a colorless oil .
OH
~C2Me ~J~H
LAH THF [~ H

To a se~ ti~n of 4-d5 ethoxyphenylacetate (1 g, 5 mmol) in dry ethyl ether at O C was added lithium aluminum hydride (380 mg, 10 mmol). After stirring at 25 C for 8 hours, the reaction mixture was poured into cold 3 M HCl. The 3 0 aqueous sol ut i rn was then extracted with ethyl ether twice. The organic layers were combined, washed with brine and dried over MgS04. Removal of the solvent gave 4-d5 ethoxyphenethyl alcohol (850 mg, 10096) as a white golid .
.

W0 95/28t.40 ' 2 1 8 7 7 9 2 ~ t - ~ ~

Hl NMR (CDCl3) 2.79 (2H, t, J= 6.6 Hz), 3.80 (2H, t, J= 6.5 Hz), 6.83 (2H, d, J= 8.6 Hz), 7.12 (2H, d, J= 8.6 Hz) .
To a solution o~ 4-d5 ethoxyphenylacetate (800 mg, 4 mmol) 5 in dry ether at 0 C was added lithium aluminum deuteride (340 mg, 8 mmol). After ~;tirring at 25 C ~or 8 hours, the reactio~ mixture was poured into cold 3 M HCl. The a~aueous soluti~n was then extracted with ethyl ether twice. The organic layers were in~d, washed with 10 brine and dried over MgSO~. Removal o~ the solvent gave 4-d5 ethoxy d2-phenethyl alcohol (680 mg, 9896) as a white solid .
EIl NMR (CDCl3) 2.79 (2H, s), 6.85 (2H, d, J= 8.5 Hz), 7.13 (2H, d, J= 8 . 5 Hz) .
Scheme 12 displays a table of 7 dif f erent tags by combination of the ~ _ ' in the horizontal column with the, r~ in the corr.osr~n~;n~ vertical column.

~w09s~640 2187792 , ,,l ~, ~

Scheme 12 hass ~ag OH
¢~ 0 3 5 OH
OH
HCD

OX
~ D
¢~ D 2 5 7 OH
Synthesis OH
COzne ~co2ne ~D
CH3CH20H ~ LAD. THF ¢~ D
OH OCH2CH3 OCHzCH3 Ol~e ~eO2C ,¢~,o OCH2CH3 PPh3. DEAD
neo2c D D
~ass=h~2 W0 9SI28640 2 1 8 7 7 ~ 2 -- t Df d-7 or d-5 deuterium 8~11m:\le8 OH OX
~H ~D
and OCD2CD3 : OCDzCD3 o d d s 7 sensitivity re5~uires that ~he samples are derivatized to form the trimethyl silyl e6ters using derivatizing agent bis (trimethylsilyl) trifluoroacetamide (;3STFA) quantitatively for 30 minutes at room temperature. The reagent and solvent are removed at room temperature under a stream of N, (g) . The samples are resolvated with ethyl acetate and injected on column for analysis using positive rh~m;r;~l ;nn;7:~t;nn ma55 ~ye~:L~ try. The d-7 and d-5 , ^~tR cannot be geparated on the GC but mass spectrometry allows the ratio of the two ~ ~,v..el~Ls present to be determined when apprry;r-tl~ly 20 ng is injected on column using t;~e ~ewlett-Packard GC/MS. The (M+NH~) ~ ion observed in Fi~ure 6 for the mass spectrum of the derivativized d-5 an~ d-7 sample show a m/z = 263 for d-7 and m/z ~ 261 for d-5 which agree with the calculated ratio. Fragments for the para-ethoYy-benzyl cation are also observed in Figure 6 at m/z = 156 for d-7 and m/z =
154 for d-5. Sensitivity may be increased to 1 ng or less injected on column if the more sensitive JEO~ SX-102 high r~Rnl1ltinn mass spectrometer were used. The JEOL SX-102 i~ about 2 orders o+ magnitude more sensitive than the ~Iewlett-Packard system. Figure 6 shows spectra and UilLI t~n~rams showing the; u~. in cl~, tography due to derivatization. Figure 5 5hows mass spectrum and cllromatogram for the underivativzed d-7 sample only.

~ w0 9~640 2 1 8 7 7 9 2 t is evident from the above description that the subject invention provides a versatile, simple method for identifying compounds, where the amount of compound present precludes any assurance of the ability to obtain ~, 5 an accurate det~rm;n~t;~n of its reaction history. The method allows for the production of extraordinarily large numbers of different products, which can be used in various screening techniques to determine biological or other activity of interest. The use of tags which are chemically inert under the process conditions allows for great v~r~at; l i ty in a variety of enviL~ q produced by the various synthetic techniques employed f or producing the products of interest. The tags can be readily synthesized and permit accurate analysis, 80 as to accurately define the nature of the composition.
A11 publications and patent applications cited in this specification are herein incorporated by reference as if each individual publication or patent application were speci~ically and individually ;nfl;r~t~d to be incorporated by reference.
Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, it will be readily Le~L to those of ordinary skill in the art in light of the t~ h; n~q of this invention that certain changes and modifications may be made thereto without departing from the spirit or scope of the appended claims.

Claims (44)

WHAT IS CLAIMED IS:

1. A compound of the formula:
F1-F2-C-E-C' wherein C-E-C' is a tag which is capable of analysis, the components of which provide for separation from other tags and allow either for detection or for separation the addition of a detectable component;
F2 is a linking component capable of being selectively cleaved to release the tag components, F2 being selected from the group consisting of , , , , , , , , , , , and ;
A is -O-, -OC(O) O-, -OC(O) -, or -NHC(O) -;
R1 is H or C1-C6 alkyl;
R5 is C1-C6 alkyl; and F1 is a functional group which allows for ready attachment of the compound to a solid support.

2. A compound of claim 1, having the formula:
F1-F2- (C (E-C')a)b wherein:
F1 is CO2H, CH2X, NR1R1, C(O)RI, OH, CHN2, SH, C(O)CHN2, S(O2)Cl, .S(O2)CHN2, N3, NO2, NO, S(O2)N3, OC(O)X, C(O)X, NCO, or NCS;
C is a bond, C1-C20 alkylene optionally substituted by 1-40 F, Cl, Br, C1-C4 alkoxy, NR4R4, OR4, or NR4, or -[(C(R4)2)m -Y-Z-Y-(C(R4)2)nY-Z-Y]p-; with the proviso that the maximum number of carbon atoms in C+C' is 20;
C' is H; F; Cl; C1-C20 alkylene optionally substituted by 1-40 F, Cl, Br, C1-C? alkoxy, NR4R4, OR4, or NR4, or -[(C(R4)2)m -Y-Z-Y-(C(R4)2)nY-Z-Y]p-; with the proviso that the maximum number of carbon atoms in C+C' is 20:
E is C1-C10 alkyl substituted by 1-20 F, Cl or Br; or Q-aryl wherein the aryl is substituted by 1-7 F, Cl, NO2, SO2R5, or substituted phenyl wherein the substituent is 1-5 F, Cl, NO2, or SO2R5;
E-C' may be -H, -OH, or amino;
R1 is H or C1-C6 alkyl;
R5 is C1-C6 alkyl;
a is 1-5;
b is 1-3;
m and n are independently 0-20;
p is 1-7;
Q is a bond, O, S, NR4, C=O, -C(O)NR5, -NR5C(O)-, -C(O)O-, or -OC(O) -;
X is a leaving group;
Y is a bond, O, S, or NHR4;
Z is a bond; phenylene optionally substituted by 1-4 F, CL, Br, C1-C6 alkyl, C1-C6 alkoxy, C1-C6 alkyl substituted by 1-13 F, Cl, or C1-C6 alkyloxy substituted by 1-13 F, Cl, or Br; (C(R4)2) 1-20; or (CF1)1-20; with the proviso that when Z is a bond one of its adjacent Y' s is also a bond and aryl is a mono- or hi-cyclic aromatic ring containing up to 10 carbon atoms and up to 2 heteroatoms selected from O, S, and N.

3. A compound of claim 2, wherein:
F1 is CO2H, OH, CHN2, C(O)CHN2, C(O)X, NCS, or CH2X;
C and C' are each independently C1-C20 alkylene unsubstituted or substituted by 1-40 F or Cl, or [O-(CH2)2-3]p;
E is C1-C10 alkyl substituted by 1-20 F or Cl; Q-aryl where aryl is a bi-cyclic aromatic ring substituted by 1-7 F or Cl; or Q-phenyl substituted by 1-5 F, Cl, NO2, or SO2R5; and Q is a bond, O, -NR5C(O) -, or -OC(O)-.

4. A compound of claim 2, having the formula:

, and wherein Ar is pentafluoro-, pentachloro-, or pentabromophenyl, 2, 3, 5, 6-tetrafluoro-4(2,3,4,5,6-pentafluorophenyl)phenyl, 2,4,6-trichlorophenyl, 2,4,5-trichlorophenyl, 2,6-dichloro-4-fluorophenyl, or 2, 3, 5, 6-tetrafluorophenyl.

5. A compound of claim 1, wherein E-C' is H, OH, or NH2.

6. The compound of claim 1, wherein F1 is CHN2, C(O)CHN2, S(O2)CHN2, N3, NO2, or SO1N3.

7. The compound of claim 2, wherein F1 is COOH, CHN2, C(O)CHN2, S(O2)CHN2, COCl, OH, SH, CH2X, or NHR1;
wherein F2 is , , , , , , , or ;
wherein A is -O- or -OC(O)O-;
?-(E-?')2 is - (CR42)1-15-(O)0-1- Ar, - (CR42)0-15- (CF?)1-15-F, -(CR42)0-15-(CF2)1-15-(CR42)0-15-H, or - (CH2)1-20-((O)0-1-(CH2)1-19)0-2?-(CH2)0-24-(O)0-1-Ar, wherein Ar is pentafluoro-, pentachloro-, or pentabromophenyl, 2,3,5,6- tetrafluoro-4(2,3,4,5,6-pentafluorophenyl)phenyl, 2,4,6-trichlorophenyl, 2,4,5-trichlorophenyl, 2,6-dichloro-4-fluorophenyl, or 2, 3, 5, 6-tetrafluorophenyl.

8. A compound having the formula:

9. A method for recording the reaction history of a reaction series on each of a plurality of unique solid supports, wherein said reaction series involves at least two stages requiring differing agents or reaction conditions resulting in a different modification as to a plurality of said unique solid supports, resulting in a plurality of different final products on different unique solid supports, employing a combination of identifiers for recording said reaction history, said identifiers each comprising a tag capable of being analyzed by mass spectroscopy so as to define the choice of agent or reaction condition and the stage in said reaction series, one or more of said tags comprising one or more deuterium atoms, said method comprising:
reacting, at a first or intermediate stage of said series, a different agent or employing a different reaction condition with each of a group of said unique solid supports, said group comprising at least one of said unique solid supports, and a combination of identifiers wherein said combination of identifiers defines the choice of agent and the stage in said series as to each group of said unique solid supports, each of said identifiers being individually bound to said unique solid support directly or through other than the tag component of a prior identifier;
mixing said groups together and then dividing said plurality of unique solid supports into a plurality of groups for a second intermediate or final stage;
and repeating said reacting at least once to provide a plurality of final products, having different products on the different individual unique solid supports.

10. A kit comprising a plurality of different separated distinguishable identifiers which differ one from another, each of which is encompassed by the formula:
F1-F2-C-E-C', each of the identifiers encoding information as to a particular choice at a particular stage in the reaction series;
wherein F1 is a functional group which allows for ready attachment of the identifier to a solid support;
F2 is a linking component capable of being selectively cleaved to release che tag;
C-E-C' is a tag which is capable of analysis, the components of which provide for separation from other tags and allow either for detection or for the addition of a detectable component;
and wherein the tag comprises one or more deuterium atoms.

11. A kit according to claim 10, wherein F2 is capable of being cleaved photochemically.

12. A kit according to claim 10, wherein F2 is capable of being cleaved oxidatively, hydrolytically, thermolytically, or reductively.

13. A library of solid supports, each solid support having a synthesized organic compound and a combination of identifiers bound thereto, each of said identifiers in each combination comprising a tag capable of being analyzed so as to define a reactant or reaction condition used to synthesize the bound organic compound, one or more of said tags comprising one or more deuterium atoms.

14. A library of solid supports according to claim 13, wherein each hound organic compound, is an oligomer which is an oligopeptide, oligonucleotide, oligosaccharide, polylipid, polyester, polyamide, polyurethane, polyurea, polyether, poly (phosphorus derivative) which is a phosphate, phosphonate, phosphoramide, phosphonamidey, phosphite, or phosphinamide, poly (sulfur derivative) which is a sulfone, sulfonate, sulfite, sulfonamide, or sulfenamide, where for the phosphorous and sulfur derivatives the indicated heteroatom for the most part will be bonded to C, H, N, O or S, or a combination thereof.

15. A library of solid supports according to claim 13, wherein each bound organic compound is a non-oligomer which is heterocyclic, aromatic, alicyclic, or aliphatic, or a combination thereof.

16. A library of solid supports according to claim 15, wherein the non-oligomer is a diazabicyclic, an azatricyclic, or a branched amide compound.

17. A library of solid supports according to claim 13, wherein each bound organic compound is linked to the support through a non-labile linkage.

18. A library of solid supports according to claim 13, wherein each bound organic compound is linked to the support through a cleavable linkage.

19. A library of solid supports according to claim 13, wherein each support is a bead of about 10-2000 µm in diameter.

20. A library of claim 13, wherein the bound organic compounds have been cleaved from the solid supports.

21. A process for identifying a compound having a characteristic of interest which comprises screening a library of solid supports according to claim 13.

22. A process according to claim 21, wherein the compounds have been cleaved from the solid supports.

23. A method for recording the reaction history of a reaction series on a solid support, wherein said reaction series involves at least two stages requiring differing agents and/or reaction conditions resulting in a different modification as to a plurality of unique solid supports, resulting in a plurality of different final products on different unique solid supports, employing a combination of identifiers for recording said reaction history, said identifiers each comprising a tag component and being characterized by defining the choice of agent or reaction condition and the stage in said series and being capable of being analyzed as to the choice and stage, said method comprising:
reacting, at a first or intermediate stage of said series, a different agent or employing a different reaction condition with each of a group of said unique solid supports, said group comprising at least one of said unique solid supports, and a combination of identifiers wherein said combination of identifiers defines the choice of agent and the stage in said series as to each group of said unique solid supports, each of said identifiers being individually bound to said unique solid support directly or through other than the tag component of a prior identifier and one or more of said identifiers comprising one or more deuterium atoms;
mixing said groups together and then dividing said plurality of unique solid supports into a plurality of groups for a second intermediate or final stage:
repeating said reacting at least once to provide a plurality of synthesized organic compounds on the different individual unique solid supports; and identifying said reaction history of at least one selected unique solid support by means of said combination of identifiers.

24. A method according to claim 23, wherein said identifying includes the stage of screening said compounds for a characteristic of interest.

25. A method for identifying the reaction history of a compound having a characteristic of interest involving a reaction series employing a method for recording the reaction history of a reaction series on each of a plurality of unique solid supports, wherein said reaction series involves at least two stages requiring differing agents and/or reaction conditions resulting in a different modification as to each of a plurality of said unique solid supports, resulting in a plurality of different synthesized organic compounds on different unique solid supports, employing combinations of identifiers for recording said reaction history, each combination of identifiers defining the choice of agent and/or reaction condition and the stage in said series, each identifier comprising a tag capable of being analyzed by mass spectroscopy as to the choice and stage, one or more of said tags comprising one or more deuterium atoms, said method comprising:
reacting, at a first or intermediate stage of said series, a different agent and/or employing a different reaction condition with each of a group of said unique solid supports, said group comprising at least one of said unique solid supports, and a combination of identifiers wherein said combination of identifiers defines the choice of agent and the stage in said series as to each group of said unique solid supports, each of said identifiers being individually bound to said unique solid supports through other than the tag component of a prior identifier by a cleavable link;
mixing said groups together and then dividing said plurality of unique solid supports into a plurality of groups for a second intermediate or final stage;
repeating said reacting to provide a plurality of compounds so that a different compound is bound to each of the different individual unique solid supports;
screening the compounds from a plurality of each of said unique solid supports for a characteristic of interest; and identifying said reaction history for at least one unique solid support having a compound having said characteristic of interest by detaching the tag members from said unique solid support and identifying said tag members by means of a differing characteristic.

26. A composition of the formula S-P1-F2-C-E-C' wherein:
S is a soluble or solid support:

C-E-C' is a tag which is capable of analysis comprising one or more deuterium atoms, the components of which tag provide for separation from other tags and allow either for detection or the addition of a detectable component:
F2 is a linking component capable of being selectively cleaved to release the tag; and F1 is a functional group which provides for attachment to the support.

27. A composition of claim 26, wherein:
S is a capillary, hollow fiber, needle, solid fiber, cellulose bead, pore-glass bead, silica gel, polystyrene bead optionally cross-linked with divinylbenzene, grafted co-poly bead, poly-acrylamide bead, latex bead, dimethylacrylamide bead optionally cross-linked with N,N'-bis-acryloyl ethylene diamine, glass particles coated with a hydrophobic polymer, or low molecular weight non-cross-linked polystyrene.

28. The method of claim 9, wherein the combination of identifiers defines a binary coding scheme.

29. The method of claim 9, wherein the identifiers are compounds of claim 1.

30. The method of claim 9, wherein the identifiers are compounds of claim 2.

31. The method of claim 9, wherein the identifiers are compounds of claim 3.

32. The method of claim 9, wherein the identifiers are compounds of claim 4.

33. The method of claim 9, further comprising detaching each tag from said unique solid supports.

34. The method of claim 33, wherein each tag is detached photochemically, oxidatively, hydrolytically, thermolytically, or reductively.

35. A method of synthesizing a chemical compound so that the structure of the compound is readily determinable, which comprises synthesizing the compound on the surface of a solid support under conditions such that the solid support at the completion of the synthesis of the compound has bound to it a plurality of identifiers which encode the reaction stages associated with the synthesis of the compound, one or more of said identifiers comprising one or more deuterium atoms.

36. A method of synthesizing a library of chemical compounds so that the structure of each compound in the library is readily determinable, the method comprising synthesizing each compound on the surface of a unique solid support under conditions such that each such unique support at the completion of the synthesis of the library of compounds has bound to it a plurality of identifiers which encode the reaction stages associated with the synthesis of the compound synthesized on such solid support, one or more of said identifiers comprising one or more deuterium atoms.

37. A method of determining the structure of a chemical compound which comprises synthesizing the compound by the method of claim 35 or 36, isolating the solid support upon which the compound was synthesized, treating the solid support so isolated so as to cause the tag components of each of the identifiers bound to the solid support to be released, determining the identity or quantity or both of each tag component so released, and deriving the structure of the compound from the identities or quantities or both of all such tag components.

38. A method of identifying a compound having a desired characteristic which comprises synthesizing a library of chemical compounds by the method of claim 36, separately testing each of the compounds in the resulting library in an assay which identifies compounds having the desired characteristic so as to identify any compounds present in the library which has the desired characteristic.

39. A method of claim 38, further comprising determining the structure of the compound so identified.

40. A library of chemical compounds, each compound in the library being bound to a unique solid support and each such solid support having bound to it a plurality of identifiers which encode the reaction stages associated with the synthesis of the compound bound to such solid support, one or more of said identifiers comprising one or more deuterium atoms.

41. A library of claim 40, wherein compounds in the library are diazabicyclic compounds, azatricyclic compounds, branched amide compounds, or peptides.

42. A method of identifying a compound having a desired characteristic which comprises testing a library of chemical compounds according to claim 39 in an assay which identifies compounds having the desired characteristic so as to identify any compound present in the library having the desired characteristic.

43. A method of claim 42, further comprising determining the structure of the compound so identified

44. A compound identified by the method of claim 42.